/// <summary>
/// This method removes sequences not having the required number of chains.
/// </summary>
/// <returns></returns>
public static List <ISequence> RemoveSequencesWithIncorrectNumberOfChains(CancellationToken cancellationToken, List <ISequence> sequenceList, int numberOfChainsRequired = 2, ProgressActionSet progressActionSet = null)
{
if (sequenceList == null || sequenceList.Count == 0)
{
throw new ArgumentOutOfRangeException(nameof(sequenceList));
}
if (progressActionSet == null)
{
throw new ArgumentNullException(nameof(progressActionSet));
}
var pdbIdListNotDistinct = FilterProteins.SequenceListToPdbIdList(sequenceList, false);
ProgressActionSet.StartAction(pdbIdListNotDistinct.Count, progressActionSet);
var workDivision = new WorkDivision <List <string> >(pdbIdListNotDistinct.Count);
for (var threadIndex = 0; threadIndex < workDivision.ThreadCount; threadIndex++)
{
var localThreadIndex = threadIndex;
Task <List <string> > task = Task.Run(() =>
{
var taskResult = pdbIdListNotDistinct.Where((a, pdbIdIndex) =>
{
if (pdbIdIndex < workDivision.ThreadFirstIndex[localThreadIndex] || pdbIdIndex > workDivision.ThreadLastIndex[localThreadIndex])
{
return(false);
}
workDivision.IncrementItemsCompleted(1);
ProgressActionSet.ProgressAction(1, progressActionSet);
ProgressActionSet.EstimatedTimeRemainingAction(workDivision.StartTicks, workDivision.ItemsCompleted, workDivision.ItemsToProcess, progressActionSet);
return(pdbIdListNotDistinct.Count(b => a == b) != numberOfChainsRequired);
}).ToList();
return(taskResult);
}, cancellationToken);
workDivision.TaskList.Add(task);
}
workDivision.WaitAllTasks();
var sequencesWithIncorrectNumberOfChains = new List <string>();
foreach (var task in workDivision.TaskList.Where(t => t != null && t.Result != null && t.IsCompleted && !t.IsCanceled && !t.IsFaulted))
{
sequencesWithIncorrectNumberOfChains.AddRange(task.Result);
}
var result = RemoveSequences(cancellationToken, sequenceList, sequencesWithIncorrectNumberOfChains);
ProgressActionSet.FinishAction(true, progressActionSet);
return(result);
}
public static List <string> RemoveStructuresWithIncorrectNumberOfChains(CancellationToken cancellationToken, string[] pdbFolders, List <string> pdbIdList = null, Dictionary <string, List <string> > pdbIdChainIdList = null, int numberChainsRequired = 2, ProgressActionSet progressActionSet = null, int totalThreads = -1)
{
if (pdbFolders == null || pdbFolders.Length == 0)
{
throw new ArgumentOutOfRangeException(nameof(pdbFolders));
}
if (pdbIdList == null || pdbIdList.Count == 0)
{
throw new ArgumentOutOfRangeException(nameof(pdbIdList));
}
if (progressActionSet == null)
{
throw new ArgumentNullException(nameof(progressActionSet));
}
var pdbFilesArray = ProteinDataBankFileOperations.GetPdbFilesArray(pdbFolders);
pdbFilesArray = ProteinDataBankFileOperations.RemoveNonWhiteListedPdbIdFromPdbFilesArray(pdbIdList, pdbFilesArray);
ProteinDataBankFileOperations.ShowMissingPdbFiles(pdbFilesArray, pdbIdList, progressActionSet);
WorkDivision <List <string> > workDivision = new WorkDivision <List <string> >(pdbFilesArray.Length, totalThreads);
ProgressActionSet.StartAction(pdbFilesArray.Length, progressActionSet);
for (int threadIndex = 0; threadIndex < workDivision.ThreadCount; threadIndex++)
{
int localThreadIndex = threadIndex;
Task <List <string> > task = Task.Run(() =>
{
var taskResult = new List <string>();
for (int pdbFileNumber = workDivision.ThreadFirstIndex[localThreadIndex]; pdbFileNumber <= workDivision.ThreadLastIndex[localThreadIndex]; pdbFileNumber++)
{
if (cancellationToken.IsCancellationRequested)
{
break;
}
try
{
string pdbFilename = pdbFilesArray[pdbFileNumber];
string proteinId = ProteinDataBankFileOperations.PdbIdFromPdbFilename(pdbFilename);
// Check if the file found is included in the white list.
if (pdbIdList != null && !pdbIdList.Contains(proteinId))
{
continue;
}
var sequenceChainIdList = pdbIdChainIdList != null ? (pdbIdChainIdList.ContainsKey(proteinId) ? pdbIdChainIdList[proteinId].ToArray() : null) : null;
int chainCount = ProteinDataBankFileOperations.PdbAtomicChainsCount(pdbFilename, sequenceChainIdList, numberChainsRequired);
if (chainCount != numberChainsRequired)
{
if (!taskResult.Contains(proteinId))
{
taskResult.Add(proteinId);
}
}
}
finally
{
workDivision.IncrementItemsCompleted(1);
ProgressActionSet.ProgressAction(1, progressActionSet);
ProgressActionSet.EstimatedTimeRemainingAction(workDivision.StartTicks, workDivision.ItemsCompleted, workDivision.ItemsToProcess, progressActionSet);
}
}
return(taskResult);
}, cancellationToken);
workDivision.TaskList.Add(task);
}
workDivision.WaitAllTasks();
ProgressActionSet.FinishAction(true, progressActionSet);
var result = new List <string>();
foreach (var task in workDivision.TaskList.Where(t => t != null && t.Result != null && t.IsCompleted && !t.IsCanceled && !t.IsFaulted))
{
result.AddRange(task.Result);
}
result = result.Distinct().ToList();
return(result);
}
/// <summary>
/// This method removes sequences from the list which are not proteins (e.g. DNA, RNA, Hybrid).
/// </summary>
/// <returns></returns>
public static List <ISequence> RemoveNonProteinAlphabetSequences(CancellationToken cancellationToken, List <ISequence> sequences, ProgressActionSet progressActionSet, int totalThreads = -1)
{
if (sequences == null || sequences.Count == 0)
{
throw new ArgumentOutOfRangeException(nameof(sequences));
}
if (progressActionSet == null)
{
throw new ArgumentNullException(nameof(progressActionSet));
}
WorkDivision <List <string> > workDivision = new WorkDivision <List <string> >(sequences.Count, totalThreads);
ProgressActionSet.StartAction(sequences.Count, progressActionSet);
for (int threadIndex = 0; threadIndex < workDivision.ThreadCount; threadIndex++)
{
int localThreadIndex = threadIndex;
Task <List <string> > task = Task.Run(() =>
{
var taskResult = new List <string>();
for (int index = workDivision.ThreadFirstIndex[localThreadIndex]; index <= workDivision.ThreadLastIndex[localThreadIndex]; index++)
{
if (cancellationToken.IsCancellationRequested)
{
break;
}
string proteinId = SequenceIdSplit.SequenceIdToPdbIdAndChainId(sequences[index].ID).PdbId;
if (sequences[index].Alphabet != Alphabets.Protein)
{
taskResult.Add(proteinId);
}
workDivision.IncrementItemsCompleted(1);
ProgressActionSet.ProgressAction(1, progressActionSet);
ProgressActionSet.EstimatedTimeRemainingAction(workDivision.StartTicks, workDivision.ItemsCompleted, workDivision.ItemsToProcess, progressActionSet);
}
return(taskResult);
}, cancellationToken);
workDivision.TaskList.Add(task);
}
workDivision.WaitAllTasks();
var result = new List <string>();
foreach (var task in workDivision.TaskList.Where(t => t != null && t.Result != null && t.IsCompleted && !t.IsCanceled && !t.IsFaulted))
{
result.AddRange(task.Result);
}
result = result.Distinct().ToList();
List <ISequence> seq = RemoveSequences(cancellationToken, sequences, result);
return(seq);
}
/// <summary>
/// This method removes specified ids from the list of sequences.
/// </summary>
/// <param name="sequences"></param>
/// <param name="sequencesToKeepOrRemove"></param>
/// <param name="options"></param>
/// <returns></returns>
public static List <ISequence> RemoveSequences(CancellationToken cancellationToken, List <ISequence> sequences, List <string> sequencesToKeepOrRemove, RemoveSequencesOptions options = RemoveSequencesOptions.RemoveSequencesInList, int totalThreads = -1)
{
if (sequences == null || sequences.Count == 0)
{
throw new ArgumentOutOfRangeException(nameof(sequences));
}
if (sequencesToKeepOrRemove == null)
{
throw new ArgumentOutOfRangeException(nameof(sequencesToKeepOrRemove));
}
if (sequencesToKeepOrRemove != null)// && sequencesToKeepOrRemove.Count > 0)
{
var workDivision = new WorkDivision <List <int> >(sequences.Count, totalThreads);
for (int threadIndex = 0; threadIndex < workDivision.ThreadCount; threadIndex++)
{
int localThreadIndex = threadIndex;
Task <List <int> > task = Task.Run(() =>
{
var taskResult = new List <int>();
for (int sequencesIndex = workDivision.ThreadFirstIndex[localThreadIndex]; sequencesIndex <= workDivision.ThreadLastIndex[localThreadIndex]; sequencesIndex++)
{
string proteinId = SequenceIdSplit.SequenceIdToPdbIdAndChainId(sequences[sequencesIndex].ID).PdbId;
if (((options == RemoveSequencesOptions.RemoveSequencesInList) && (sequencesToKeepOrRemove.Contains(proteinId))) ||
((options == RemoveSequencesOptions.RemoveSequencesNotInList) && (!sequencesToKeepOrRemove.Contains(proteinId))))
{
taskResult.Add(sequencesIndex);
}
workDivision.IncrementItemsCompleted(1);
}
return(taskResult);
}, cancellationToken);
workDivision.TaskList.Add(task);
}
workDivision.WaitAllTasks();
var sequenceIndexesToRemove = new List <int>();
foreach (var task in workDivision.TaskList.Where(t => t != null && t.Result != null && t.IsCompleted && !t.IsCanceled && !t.IsFaulted))
{
sequenceIndexesToRemove.AddRange(task.Result);
}
sequenceIndexesToRemove = sequenceIndexesToRemove.Distinct().ToList();
sequenceIndexesToRemove.Sort();
for (int sequenceIndexesToRemoveIndex = sequenceIndexesToRemove.Count - 1; sequenceIndexesToRemoveIndex >= 0; sequenceIndexesToRemoveIndex--)
{
sequences.RemoveAt(sequenceIndexesToRemove[sequenceIndexesToRemoveIndex]);
}
}
return(sequences);
}
public static List <ISequence> RemoveDuplicates(CancellationToken cancellationToken, List <ISequence> sequences, ProgressActionSet progressActionSet, int totalThreads = -1)
{
if (sequences == null || sequences.Count == 0)
{
throw new ArgumentOutOfRangeException(nameof(sequences));
}
if (progressActionSet == null)
{
throw new ArgumentNullException(nameof(progressActionSet));
}
var pdbIdList = SequenceListToPdbIdList(sequences);
var pdbIdSequences = pdbIdList.Select(a => sequences.Where(b => SequenceIdSplit.SequenceIdToPdbIdAndChainId(b.ID).PdbId == a).ToList()).ToList();
var workDivision = new WorkDivision(pdbIdList.Count, totalThreads);
ProgressActionSet.StartAction(pdbIdList.Count, progressActionSet);
var done = new List <ISequence>();
var remove = new List <ISequence>();
var removeLock = new object();
for (int threadIndex = 0; threadIndex < workDivision.ThreadCount; threadIndex++)
{
int localThreadIndex = threadIndex;
var task = Task.Run(() =>
{
for (int index = workDivision.ThreadFirstIndex[localThreadIndex]; index <= workDivision.ThreadLastIndex[localThreadIndex]; index++)
{
if (cancellationToken.IsCancellationRequested)
{
break;
}
var iterationPdbId = pdbIdList[index];
var iterationPdbIdSeqs = pdbIdSequences[index];// sequences.Where(a => SequenceIdSplit.SequenceIdToPdbIdAndChainId(a.ID).PdbId == pdbId).ToList();
//var seq = sequences[index];
//var seqid = SequenceIdSplit.SequenceIdToPdbIdAndChainId(seq.ID).PdbId.ToUpperInvariant();
lock (done)
{
if (iterationPdbIdSeqs.All(done.Contains))
{
continue;
}
}
foreach (var pdbIdSeqSet in pdbIdSequences)
{
if (pdbIdSeqSet == iterationPdbIdSeqs)
{
continue;
}
foreach (var pdbIdSeq in pdbIdSeqSet)
{
foreach (var iterationPdbIdSeq in iterationPdbIdSeqs)
{
}
}
}
// find sequences equal to the current iteration item
//var equalseq = sequences.Where(a => a.SequenceEqual(seq)).ToList();
/*
* var equalseq = sequences.Where(a => AlignedSequenceSimilarityPercentage(seq,a) >= 90).ToList();
*
*
*
* // get a list of pdbids, ordered alphabetically
* var equalseqids = equalseq.Select(p => SequenceIdSplit.SequenceIdToPdbIdAndChainId(p.ID).PdbId.ToUpperInvariant()).OrderBy(p => p).ToList();
*
* // one or more of the chains might have a difference sequence and so not in the list, update by the ids in the list
* //equalseq = sequences.Where(p => equalseqids.Contains(SequenceIdSplit.SequenceIdToPdbIdAndChainId(p.ID).PdbId.ToUpperInvariant())).ToList();
*
* // add this iteration item and all others with the same sequence to a list to skip in future
* lock (done)
* {
* done.AddRange(equalseq);
* }
*
* // keep the very last item in the list and all with the same pdbid that it has
* var keepid = equalseqids.Last();
* var equalseqkeep = equalseq.Where(p => SequenceIdSplit.SequenceIdToPdbIdAndChainId(p.ID).PdbId.ToUpperInvariant() == keepid).ToList();
*
* // remove the sequences to keep from the removal list
* equalseq = equalseq.Where(a => !equalseqkeep.Contains(a)).ToList();
*
* lock (remove)
* {
* remove.AddRange(equalseq);
* }
*/
workDivision.IncrementItemsCompleted(1);
ProgressActionSet.ProgressAction(1, progressActionSet);
ProgressActionSet.EstimatedTimeRemainingAction(workDivision.StartTicks, workDivision.ItemsCompleted, workDivision.ItemsToProcess, progressActionSet);
}
}, cancellationToken);
workDivision.TaskList.Add(task);
}
workDivision.WaitAllTasks();
var remove2 = remove.Distinct().ToList();
return(RemoveSequences(cancellationToken, sequences, remove2.Select(p => SequenceIdSplit.SequenceIdToPdbIdAndChainId(p.ID).PdbId.ToUpperInvariant()).ToList()));
}
/// <summary>
/// Perform sequence alignment on the chains of each protein to see if it is a homodimer or heterodimer
/// </summary>
/// <returns></returns>
public static DimerSequenceTypeCategories <string> SplitDimersHomoHetero(CancellationToken cancellationToken, List <ISequence> sequences, decimal minimumHeterodimerSimilarityRequired = 30.0m, decimal minimumHomodimerSimiliarityRequired = 90.0m, ProgressActionSet progressActionSet = null, int totalThreads = -1)
{
if (sequences == null || sequences.Count == 0)
{
throw new ArgumentOutOfRangeException(nameof(sequences));
}
if (progressActionSet == null)
{
throw new ArgumentNullException(nameof(progressActionSet));
}
var workDivision = new WorkDivision <DimerSequenceTypeCategories <string> >(sequences.Count, totalThreads);
ProgressActionSet.StartAction(sequences.Count, progressActionSet);
for (int threadIndex = 0; threadIndex < workDivision.ThreadCount; threadIndex++)
{
int localThreadIndex = threadIndex;
Task <DimerSequenceTypeCategories <string> > task = Task.Run(() =>
{
var taskResult = new DimerSequenceTypeCategories <string>();
for (int sequencesIndexA = workDivision.ThreadFirstIndex[localThreadIndex]; sequencesIndexA <= workDivision.ThreadLastIndex[localThreadIndex]; sequencesIndexA++)
{
if (cancellationToken.IsCancellationRequested)
{
break;
}
string proteinIdA = SequenceIdSplit.SequenceIdToPdbIdAndChainId(sequences[sequencesIndexA].ID).PdbId;
for (int sequencesIndexB = 0; sequencesIndexB < sequences.Count; sequencesIndexB++)
{
if (cancellationToken.IsCancellationRequested)
{
break;
}
// Don't align the same sequence index. Skip calculating indexes already calculated. Perform alignment operation if protein id is the same.
var proteinIdB = SequenceIdSplit.SequenceIdToPdbIdAndChainId(sequences[sequencesIndexB].ID).PdbId;
if (sequencesIndexA == sequencesIndexB || sequencesIndexB < sequencesIndexA || proteinIdA != proteinIdB)
{
continue;
}
var dimerType = FindDimerType(sequences[sequencesIndexA], sequences[sequencesIndexB], minimumHeterodimerSimilarityRequired, minimumHomodimerSimiliarityRequired);
if (dimerType == DimerType.HeteroDimer)
{
taskResult.HeteroDimerPdbIdList.Add(proteinIdA);
}
else if (dimerType == DimerType.HomoDimer)
{
taskResult.HomoDimerPdbIdList.Add(proteinIdA);
}
else if (dimerType == DimerType.HomologyDimer)
{
taskResult.HomologyDimerPdbIdList.Add(proteinIdA);
}
}
workDivision.IncrementItemsCompleted(1);
ProgressActionSet.ProgressAction(1, progressActionSet);
ProgressActionSet.EstimatedTimeRemainingAction(workDivision.StartTicks, workDivision.ItemsCompleted, workDivision.ItemsToProcess, progressActionSet);
}
return(taskResult);
}, cancellationToken);
workDivision.TaskList.Add(task);
}
workDivision.WaitAllTasks();
var dimerSequenceTypeCategories = new DimerSequenceTypeCategories <string>();
foreach (var task in workDivision.TaskList.Where(t => t != null && t.IsCompleted && !t.IsFaulted && !t.IsCanceled && t.Result != null))
{
dimerSequenceTypeCategories.HeteroDimerPdbIdList.AddRange(task.Result.HeteroDimerPdbIdList);
dimerSequenceTypeCategories.HomoDimerPdbIdList.AddRange(task.Result.HomoDimerPdbIdList);
dimerSequenceTypeCategories.HomologyDimerPdbIdList.AddRange(task.Result.HomologyDimerPdbIdList);
}
dimerSequenceTypeCategories.HeteroDimerPdbIdList = dimerSequenceTypeCategories.HeteroDimerPdbIdList.Distinct().ToList();
dimerSequenceTypeCategories.HomoDimerPdbIdList = dimerSequenceTypeCategories.HomoDimerPdbIdList.Distinct().ToList();
dimerSequenceTypeCategories.HomologyDimerPdbIdList = dimerSequenceTypeCategories.HomologyDimerPdbIdList.Distinct().ToList();
return(dimerSequenceTypeCategories);
}
/// <summary>
/// Generate stats of interactions... also removes proteins not meeting minimum interactions requirement.
/// </summary>
/// <param name="cancellationToken"></param>
/// <param name="pdbFolders"></param>
/// <param name="pdbIdList"></param>
/// <param name="progressBar"></param>
/// <param name="estimatedTimeRemainingLabel"></param>
/// <returns></returns>
public static Dictionary <string, decimal> CalculateStructureSymmetry(CancellationToken cancellationToken, decimal maxAtomInterationDistance, string[] pdbFolders, List <string> pdbIdList = null, Dictionary <string, List <string> > pdbIdChainIdList = null, ProgressActionSet progressActionSet = null, int totalThreads = -1)
{
if (pdbFolders == null || pdbFolders.Length == 0)
{
throw new ArgumentOutOfRangeException(nameof(pdbFolders));
}
if (pdbIdList == null || pdbIdList.Count == 0)
{
throw new ArgumentOutOfRangeException(nameof(pdbIdList));
}
if (progressActionSet == null)
{
throw new ArgumentOutOfRangeException(nameof(progressActionSet));
}
const int requiredNumberOfChains = 2;
string[] pdbFilesArray = ProteinDataBankFileOperations.RemoveNonWhiteListedPdbIdFromPdbFilesArray(pdbIdList, ProteinDataBankFileOperations.GetPdbFilesArray(pdbFolders));
WorkDivision <Dictionary <string, decimal> > workDivision = new WorkDivision <Dictionary <string, decimal> >(pdbFilesArray.Length, totalThreads);
ProteinDataBankFileOperations.ShowMissingPdbFiles(pdbFilesArray, pdbIdList, progressActionSet);
ProgressActionSet.StartAction(pdbFilesArray.Length, progressActionSet);
for (int threadIndex = 0; threadIndex < workDivision.ThreadCount; threadIndex++)
{
int localThreadIndex = threadIndex;
Task <Dictionary <string, decimal> > task = Task.Run(() =>
{
var taskResult = new Dictionary <string, decimal>();
for (int pdbFileNumber = workDivision.ThreadFirstIndex[localThreadIndex]; pdbFileNumber <= workDivision.ThreadLastIndex[localThreadIndex]; pdbFileNumber++)
{
if (cancellationToken.IsCancellationRequested)
{
break;
}
try
{
string pdbFilename = pdbFilesArray[pdbFileNumber];
string proteinId = ProteinDataBankFileOperations.PdbIdFromPdbFilename(pdbFilename);
// Check if the file found is included in the white list.
if (pdbIdList != null && !pdbIdList.Contains(proteinId))
{
continue;
}
var chainIdList = pdbIdChainIdList != null ? (proteinId != null && pdbIdChainIdList.ContainsKey(proteinId) ? pdbIdChainIdList[proteinId].ToArray() : null) : null;
// Get atom chains.
ProteinChainListContainer proteinFileChains = ProteinDataBankFileOperations.PdbAtomicChains(pdbFilename, chainIdList, requiredNumberOfChains, requiredNumberOfChains, true);
if (proteinFileChains == null || proteinFileChains.ChainList == null || proteinFileChains.ChainList.Count != 2 ||
proteinFileChains.ChainList[StaticValues.ChainA] == null || proteinFileChains.ChainList[StaticValues.ChainA].AtomList == null || proteinFileChains.ChainList[StaticValues.ChainA].AtomList.Count == 0 ||
proteinFileChains.ChainList[StaticValues.ChainB] == null || proteinFileChains.ChainList[StaticValues.ChainB].AtomList == null || proteinFileChains.ChainList[StaticValues.ChainB].AtomList.Count == 0)
{
continue;
}
// Make a list to save interactions found.
var interactionMatchPercentage = new InteractionMatchPercentage(proteinId);
List <AtomPair> interactions = SearchInteractions.FindInteractions(cancellationToken, maxAtomInterationDistance, pdbFilename, pdbIdChainIdList);
interactionMatchPercentage.IncrementTotalInteractions(interactions.Count);
for (int interactionsIndex = 0; interactionsIndex < interactions.Count; interactionsIndex++)
{
interactionMatchPercentage.AddResidueSequenceIndex(StaticValues.ChainA, interactions[interactionsIndex].Atom1.resSeq.FieldValue);
interactionMatchPercentage.AddResidueSequenceIndex(StaticValues.ChainB, interactions[interactionsIndex].Atom2.resSeq.FieldValue);
}
InteractionMatchPercentage.CalculatePercentageResult calculatedPercentage = interactionMatchPercentage.CalculatePercentage();
taskResult.Add(interactionMatchPercentage.ProteinId, calculatedPercentage.InteractionMatchPercentageAverage);
}
finally
{
workDivision.IncrementItemsCompleted(1);
ProgressActionSet.ProgressAction(1, progressActionSet);
ProgressActionSet.EstimatedTimeRemainingAction(workDivision.StartTicks, workDivision.ItemsCompleted, workDivision.ItemsToProcess, progressActionSet);
}
}
return(taskResult);
}, cancellationToken);
workDivision.TaskList.Add(task);
}
workDivision.WaitAllTasks();
var result = new Dictionary <string, decimal>();
foreach (var task in workDivision.TaskList.Where(t => t != null && t.Result != null && t.IsCompleted && !t.IsCanceled && !t.IsFaulted))
{
foreach (var kvp in task.Result)
{
//if (result.ContainsKey(kvp.Key))
//{
// Console.WriteLine("Key already exists: '" + kvp.Key + "'");
//}
result.Add(kvp.Key, kvp.Value);
}
}
return(result);
}
/// <summary>
/// Makes spreadsheets with scientific data outputs about given proteins.
/// </summary>
/// <param name="cancellationToken"></param>
/// <param name="pdbFolders">The location of the PDB files</param>
/// <param name="pdbIdList">The PDB files which should be used.</param>
/// <param name="consoleTextBox"></param>
/// <param name="progressBar">User proteinInterface progress bar for user feedback.</param>
/// <param name="estimatedTimeRemainingLabel">User proteinInterface estimated time remaining label for user feedback.</param>
/// <param name="requestedTotalThreads"></param>
/// <returns>Returns the generated spreadsheets with scientific data.</returns>
public static List <List <SpreadsheetCell[]> > MakeHomodimerStatisticsSpreadsheetsList(CancellationToken cancellationToken, decimal maxAtomInterationDistance, string[] pdbFolders, List <string> pdbIdList = null, Dictionary <string, List <string> > pdbIdChainIdList = null, ProgressActionSet progressActionSet = null, int requestedTotalThreads = -1)
{
if (pdbFolders == null || pdbFolders.Length == 0)
{
throw new ArgumentOutOfRangeException(nameof(pdbFolders));
}
if (pdbIdList == null || pdbIdList.Count == 0)
{
throw new ArgumentOutOfRangeException(nameof(pdbIdList));
}
if (progressActionSet == null)
{
throw new ArgumentNullException(nameof(progressActionSet));
}
// this method creates
// 1. a list of interactions
// 2. a list of symmetry percentage
// 3. an "expected" heatmap by combining every possible a/b amino acid combination
// 4. an actual heatmap for the proteinInterfaces
// 5. normalised versions of both of the heatmaps
string[] pdbFilesArray = ProteinDataBankFileOperations.RemoveNonWhiteListedPdbIdFromPdbFilesArray(pdbIdList, ProteinDataBankFileOperations.GetPdbFilesArray(pdbFolders));
//var interactionRecordList = new List<ProteinInteractionRecord>();
//var interactionMatchPercentageList = new List<InteractionMatchPercentage>();
//var wholeProteinChainsAminoAcidCounter = new List<AminoAcidChainComposition>();
//var interactionChainsAminoAcidCounter = new List<AminoAcidChainComposition>();
//var interactionsAminoAcidToAminoAcidCounter = new AminoAcidPairCompositionMatrix();
////var wholeProteinAminoAcidToAminoAcidCounter2x2 = new AminoAcidPairCompositionMatrix(); // composition of every amino acid paired in every possible combination
var workDivision = new WorkDivision <HomodimersStatisticsMinerTaskResult>(pdbFilesArray.Length, requestedTotalThreads);
ProgressActionSet.StartAction(pdbFilesArray.Length, progressActionSet);
int checkAllFilesProcessed = 0;
var lockCheckAllFilesProcessed = new object();
var pdbFilesProcessed = new bool[pdbFilesArray.Length];
Array.Clear(pdbFilesProcessed, 0, pdbFilesProcessed.Length);
for (int threadIndex = 0; threadIndex < workDivision.ThreadCount; threadIndex++)
{
int localThreadIndex = threadIndex;
Task <HomodimersStatisticsMinerTaskResult> task = Task.Run(() =>
{
var result = new HomodimersStatisticsMinerTaskResult();
for (int pdbFileNumber = workDivision.ThreadFirstIndex[localThreadIndex]; pdbFileNumber <= workDivision.ThreadLastIndex[localThreadIndex]; pdbFileNumber++)
{
if (cancellationToken.IsCancellationRequested)
{
break;
}
lock (lockCheckAllFilesProcessed)
{
checkAllFilesProcessed++;
pdbFilesProcessed[pdbFileNumber] = true;
}
try
{
string pdbFilename = pdbFilesArray[pdbFileNumber];
string proteinId = ProteinDataBankFileOperations.PdbIdFromPdbFilename(pdbFilename);
// Check if the file found is included in the white list.
if (/*pdbIdList != null && */ !pdbIdList.Contains(proteinId))
{
ProgressActionSet.Report("Error: " + proteinId + " was not in the PDB ID white list.", progressActionSet);
continue;
}
List <AtomPair> interactions = SearchInteractions.FindInteractions(cancellationToken, maxAtomInterationDistance, pdbFilename, pdbIdChainIdList);
// Make a list to save interactions found.
var interactionMatchPercentage = new InteractionMatchPercentage(proteinId);
var chainAminoAcidCounterA1X1 = new AminoAcidChainComposition(proteinId, "A");
var chainAminoAcidCounterB1X1 = new AminoAcidChainComposition(proteinId, "B");
var chainInteractionAminoAcidCounterA = new AminoAcidChainComposition(proteinId, "A");
var chainInteractionAminoAcidCounterB = new AminoAcidChainComposition(proteinId, "B");
if (interactions != null && interactions.Count > 0)
{
interactionMatchPercentage.IncrementTotalInteractions(interactions.Count);
for (int interactionsIndex = 0; interactionsIndex < interactions.Count; interactionsIndex++)
{
chainInteractionAminoAcidCounterA.IncrementAminoAcidCount(interactions[interactionsIndex].Atom1.resName.FieldValue);
chainInteractionAminoAcidCounterB.IncrementAminoAcidCount(interactions[interactionsIndex].Atom2.resName.FieldValue);
result.InteractionRecordList.Add(new ProteinInteractionRecord(proteinId, interactionsIndex + 1, interactions[interactionsIndex]));
interactionMatchPercentage.AddResidueSequenceIndex(StaticValues.ChainA, interactions[interactionsIndex].Atom1.resSeq.FieldValue);
interactionMatchPercentage.AddResidueSequenceIndex(StaticValues.ChainB, interactions[interactionsIndex].Atom2.resSeq.FieldValue);
result.InteractionsAminoAcidToAminoAcidCounter.IncrementAminoAcidCount(interactions[interactionsIndex].Atom1.resName.FieldValue, interactions[interactionsIndex].Atom2.resName.FieldValue);
}
}
var chainIdList = pdbIdChainIdList != null ? (pdbIdChainIdList.ContainsKey(proteinId) ? pdbIdChainIdList[proteinId].ToArray() : null) : null;
ProteinChainListContainer proteinFileChains = ProteinDataBankFileOperations.PdbAtomicChains(pdbFilename, chainIdList, 2, 2, true);
if (proteinFileChains == null || proteinFileChains.ChainList == null || proteinFileChains.ChainList.Count != 2 ||
proteinFileChains.ChainList[StaticValues.ChainA] == null || proteinFileChains.ChainList[StaticValues.ChainA].AtomList == null || proteinFileChains.ChainList[StaticValues.ChainA].AtomList.Count == 0 ||
proteinFileChains.ChainList[StaticValues.ChainB] == null || proteinFileChains.ChainList[StaticValues.ChainB].AtomList == null || proteinFileChains.ChainList[StaticValues.ChainB].AtomList.Count == 0)
{
if (!File.Exists(pdbFilename))
{
ProgressActionSet.Report("Error: " + pdbFilename + " (" + proteinId + ") file not found", progressActionSet);
}
else
{
int proteinFileChainCount = ProteinDataBankFileOperations.PdbAtomicChainsCount(pdbFilename);
ProgressActionSet.Report("Error: " + proteinId + " did not have exactly 2 chains (" + proteinFileChainCount + " chains found) - skipping.", progressActionSet);
}
continue;
}
// count total of how many of each type of amino acids are in Chain A.
for (int atomIndexA = 0; atomIndexA < proteinFileChains.ChainList[StaticValues.ChainA].AtomList.Count; atomIndexA++)
{
chainAminoAcidCounterA1X1.IncrementAminoAcidCount(proteinFileChains.ChainList[StaticValues.ChainA].AtomList[atomIndexA].resName.FieldValue);
}
// count total of how many of each type of amino acids are in Chain B.
for (int atomIndexB = 0; atomIndexB < proteinFileChains.ChainList[StaticValues.ChainB].AtomList.Count; atomIndexB++)
{
chainAminoAcidCounterB1X1.IncrementAminoAcidCount(proteinFileChains.ChainList[StaticValues.ChainB].AtomList[atomIndexB].resName.FieldValue);
}
interactionMatchPercentage.CalculatePercentage();
result.InteractionMatchPercentageList.Add(interactionMatchPercentage);
result.WholeProteinChainsAminoAcidCounter.Add(chainAminoAcidCounterA1X1);
result.WholeProteinChainsAminoAcidCounter.Add(chainAminoAcidCounterB1X1);
result.InteractionChainsAminoAcidCounter.Add(chainInteractionAminoAcidCounterA);
result.InteractionChainsAminoAcidCounter.Add(chainInteractionAminoAcidCounterB);
}
finally
{
workDivision.IncrementItemsCompleted(1);
ProgressActionSet.ProgressAction(1, progressActionSet);
ProgressActionSet.EstimatedTimeRemainingAction(workDivision.StartTicks, workDivision.ItemsCompleted, workDivision.ItemsToProcess, progressActionSet);
}
}
return(result);
}, cancellationToken);
workDivision.TaskList.Add(task);
}
workDivision.WaitAllTasks();
ProgressActionSet.FinishAction(true, progressActionSet);
// merge all instances of the results
var spreadsheetTaskResult = new HomodimersStatisticsMinerTaskResult();
foreach (var task in workDivision.TaskList.Where(t => t != null && t.IsCompleted && !t.IsCanceled && !t.IsFaulted && t.Result != null))
{
if (task.Result.InteractionChainsAminoAcidCounter != null && task.Result.InteractionChainsAminoAcidCounter.Count > 0)
{
spreadsheetTaskResult.InteractionChainsAminoAcidCounter.AddRange(task.Result.InteractionChainsAminoAcidCounter);
}
if (task.Result.InteractionMatchPercentageList != null && task.Result.InteractionMatchPercentageList.Count > 0)
{
spreadsheetTaskResult.InteractionMatchPercentageList.AddRange(task.Result.InteractionMatchPercentageList);
}
if (task.Result.InteractionRecordList != null && task.Result.InteractionRecordList.Count > 0)
{
spreadsheetTaskResult.InteractionRecordList.AddRange(task.Result.InteractionRecordList);
}
if (task.Result.InteractionsAminoAcidToAminoAcidCounter != null)
{
if (task.Result.InteractionsAminoAcidToAminoAcidCounter.AminoAcidToAminoAcid != null)
{
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
var totalGroups = AminoAcidGroups.AminoAcidGroups.GetTotalSubgroups(enumAminoAcidGroups);
for (int x = 0; x < totalGroups; x++)
{
for (int y = 0; y < totalGroups; y++)
{
spreadsheetTaskResult.InteractionsAminoAcidToAminoAcidCounter.AminoAcidToAminoAcid[(int)enumAminoAcidGroups][x, y] +=
task.Result.InteractionsAminoAcidToAminoAcidCounter.AminoAcidToAminoAcid[(int)enumAminoAcidGroups][x, y];
}
}
}
}
}
if (task.Result.WholeProteinChainsAminoAcidCounter != null && task.Result.WholeProteinChainsAminoAcidCounter.Count > 0)
{
spreadsheetTaskResult.WholeProteinChainsAminoAcidCounter.AddRange(task.Result.WholeProteinChainsAminoAcidCounter);
}
}
if (pdbFilesProcessed.Count(file => file == false) > 0)
{
ProgressActionSet.Report("ERROR: " + pdbFilesProcessed.Count(file => file == false) + " PDB FILES WERE SKIPPED! 0x01", progressActionSet);
}
else
{
ProgressActionSet.Report("CHECK: NO PDB FILES WERE SKIPPED! 0x01", progressActionSet);
}
if (checkAllFilesProcessed != pdbFilesArray.Length)
{
ProgressActionSet.Report("ERROR: " + (pdbFilesArray.Length - checkAllFilesProcessed) + " PDB FILES WERE SKIPPED! 0x02", progressActionSet);
}
else
{
ProgressActionSet.Report("CHECK: NO PDB FILES WERE SKIPPED! 0x02", progressActionSet);
}
spreadsheetTaskResult.WholeProteinChainsAminoAcidCounter = spreadsheetTaskResult.WholeProteinChainsAminoAcidCounter.OrderBy(a => a.ProteinId).ThenBy(b => b.ChainId).ToList();
spreadsheetTaskResult.InteractionChainsAminoAcidCounter = spreadsheetTaskResult.InteractionChainsAminoAcidCounter.OrderBy(a => a.ProteinId).ThenBy(b => b.ChainId).ToList();
AminoAcidChainComposition wholeProteinChainsTotals = AminoAcidChainComposition.TotalFromAminoAcidChainCompositionList(spreadsheetTaskResult.WholeProteinChainsAminoAcidCounter);
AminoAcidChainComposition interactionChainsTotals = AminoAcidChainComposition.TotalFromAminoAcidChainCompositionList(spreadsheetTaskResult.InteractionChainsAminoAcidCounter);
AminoAcidPairCompositionMatrix wholeProteinAminoAcidToAminoAcidCounter1X1 = AminoAcidChainComposition.ConvertToMatrix(wholeProteinChainsTotals);
var results = new List <List <SpreadsheetCell[]> >();
{
/* start test */
var spreadsheet1 = new List <SpreadsheetCell[]>();
spreadsheet1.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% TEST SHEET 0"), });
spreadsheet1.Add(new[] { new SpreadsheetCell("TEST SHEET 0"), });
foreach (AminoAcidChainComposition item in spreadsheetTaskResult.WholeProteinChainsAminoAcidCounter)
{
//spreadsheet1.Add(item.ProteinId);
//spreadsheet1.Add(item.ChainId);
spreadsheet1.Add(item.SpreadsheetDataRow());
}
results.Add(spreadsheet1);
spreadsheet1 = null;
/* end test */
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
{
var spreadsheet2 = new List <SpreadsheetCell[]>();
spreadsheet2.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% L Interaction Symmetry"), });
spreadsheet2.Add(new[] { new SpreadsheetCell("Homodimers - List - Interaction Count And Interaction Match Percentage (Symmetry Measurement)") });
spreadsheet2.Add(InteractionMatchPercentage.SpreadsheetColumnHeadersRow());
var range2 = spreadsheetTaskResult.InteractionMatchPercentageList.Select(record => record.SpreadsheetDataRow()).ToList();
//range2.Sort();
range2 = range2
.OrderBy(a => a[0].CellData)
.ThenBy(a => a[1].CellData)
.ThenBy(a => a[2].CellData)
.ThenBy(a => a[3].CellData)
.ThenBy(a => a[4].CellData)
.ThenBy(a => a[5].CellData)
.ThenBy(a => a[6].CellData)
.ThenBy(a => a[7].CellData)
.ThenBy(a => a[8].CellData)
.ToList();
spreadsheet2.AddRange(range2);
range2 = null;
results.Add(spreadsheet2);
var spreadsheetHistogram2 = new List <SpreadsheetCell[]>();
spreadsheetHistogram2.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% HG Interaction Symmetry"), });
spreadsheetHistogram2.Add(new[] { new SpreadsheetCell("Homodimers - List - Interaction Count And Interaction Match Percentage (Symmetry Measurement) Histogram") });
spreadsheetHistogram2.AddRange(Histogram.MatrixToHistogram(spreadsheet2.ToArray(), Histogram.MakeBinDecimals(0, 100, 9, 10), new[] { 6, 7, 8 }, 2, -1, true));
results.Add(spreadsheetHistogram2);
spreadsheet2 = null;
spreadsheetHistogram2 = null;
}
//
{
var spreadsheet3 = new List <SpreadsheetCell[]>();
spreadsheet3.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% L Interaction Records"), });
spreadsheet3.Add(new[] { new SpreadsheetCell("Homodimers - List - Protein Interaction Record"), });
spreadsheet3.Add(ProteinInteractionRecord.TsvColumnHeadersRow());
var range3 = spreadsheetTaskResult.InteractionRecordList.Select(record => record.SpreadsheetDataRow()).ToList();
//range3.Sort();
range3 = range3
.OrderBy(a => a[0].CellData)
.ThenBy(a => a[1].CellData)
.ThenBy(a => a[3].CellData)
.ThenBy(a => a[5].CellData)
.ThenBy(a => a[13].CellData)
.ThenBy(a => a[15].CellData)
.ToList();
spreadsheet3.AddRange(range3);
range3 = null;
results.Add(spreadsheet3);
var spreadsheetHistogram3 = new List <SpreadsheetCell[]>();
spreadsheetHistogram3.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% L Interaction Records Histogram"), });
spreadsheetHistogram3.Add(new[] { new SpreadsheetCell("Homodimers - List - Protein Interaction Record - Histogram"), });
spreadsheetHistogram3.AddRange(Histogram.MatrixToHistogram(spreadsheet3.ToArray(), Histogram.MakeBinDecimals(0m, 5m, 0m, 0.05m), new[] { 1 }, 2, -1, true));
results.Add(spreadsheetHistogram3);
//spreadsheet3 = Histogram.InsertMatrixOverwrite(spreadsheet3.ToArray(), histogram3, 2, Histogram.MaxColumns(spreadsheet3.ToArray()) + 1).ToList();
spreadsheet3 = null;
spreadsheetHistogram3 = null;
}
//
{
var spreadsheet4 = new List <SpreadsheetCell[]>();
spreadsheet4.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% L Interaction Count - A-Z"), });
spreadsheet4.Add(new[] { new SpreadsheetCell("Homodimers - List - Protein Amino Acid Count - Interactions - A to Z"), });
spreadsheet4.Add(AminoAcidChainComposition.SpreadsheetTitleRow());
var range4 = spreadsheetTaskResult.InteractionChainsAminoAcidCounter.Select(record => record.SpreadsheetDataRow()).ToList();
//range4.Sort();
range4 = range4
.OrderBy(a => a[0].CellData)
.ThenBy(a => a[1].CellData)
.ToList();
spreadsheet4.AddRange(range4);
range4 = null;
spreadsheet4.Add(interactionChainsTotals.SpreadsheetDataRow());
results.Add(spreadsheet4);
spreadsheet4 = null;
}
//
{
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
var spreadsheet5 = new List <SpreadsheetCell[]>();
spreadsheet5.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% L Interaction Count - Groups " + enumAminoAcidGroups), });
spreadsheet5.Add(new[] { new SpreadsheetCell("Homodimers - List - Protein Amino Acid Count - Interactions - Acid Groups " + enumAminoAcidGroups), });
spreadsheet5.Add(AminoAcidChainComposition.SpreadsheetGroupsTitleRow(enumAminoAcidGroups));
var range5 = spreadsheetTaskResult.InteractionChainsAminoAcidCounter.Select(record => record.SpreadsheetGroupsDataRow(enumAminoAcidGroups)).ToList();
//range4.Sort();
range5 = range5
.OrderBy(a => a[0].CellData)
.ThenBy(a => a[1].CellData)
.ToList();
spreadsheet5.AddRange(range5);
range5 = null;
spreadsheet5.Add(interactionChainsTotals.SpreadsheetGroupsDataRow(enumAminoAcidGroups));
results.Add(spreadsheet5);
spreadsheet5 = null;
}
}
//
{
var spreadsheet6 = new List <SpreadsheetCell[]>();
spreadsheet6.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% L Entire Count - A-Z"), });
spreadsheet6.Add(new[] { new SpreadsheetCell("Homodimers - List - Protein Amino Acid Count - All Atoms - A to Z"), });
spreadsheet6.Add(AminoAcidChainComposition.SpreadsheetTitleRow());
var range6 = spreadsheetTaskResult.WholeProteinChainsAminoAcidCounter.Select(record => record.SpreadsheetDataRow()).ToList();
//range6.Sort();
range6 = range6
.OrderBy(a => a[0].CellData)
.ThenBy(a => a[1].CellData)
.ToList();
spreadsheet6.AddRange(range6);
range6 = null;
spreadsheet6.Add(wholeProteinChainsTotals.SpreadsheetDataRow());
results.Add(spreadsheet6);
var spreadsheetHistogram6 = new List <SpreadsheetCell[]>();
spreadsheetHistogram6.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% L Entire Count - A-Z - Historgram"), });
spreadsheetHistogram6.Add(new[] { new SpreadsheetCell("Homodimers - List - Protein Amino Acid Count - All Atoms - A to Z - Histogram"), });
spreadsheetHistogram6.AddRange(Histogram.MatrixToHistogram(spreadsheet6.ToArray(), Histogram.MakeBinDecimals(0, 10500, 0, 500), new[] { 28 }, 2, -1, true));
spreadsheetHistogram6.Add(new [] { new SpreadsheetCell(""), });
spreadsheetHistogram6.AddRange(Histogram.MatrixToHistogram(spreadsheet6.ToArray(), Histogram.MakeBinDecimals(0, 1000, 0, 100), new[] { 28 }, 2, -1, true));
results.Add(spreadsheetHistogram6);
spreadsheet6 = null;
spreadsheetHistogram6 = null;
}
//
{
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
var spreadsheet7 = new List <SpreadsheetCell[]>();
spreadsheet7.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% L Entire Count - Groups " + enumAminoAcidGroups), });
spreadsheet7.Add(new[] { new SpreadsheetCell("Homodimers - List - Protein Amino Acid Count - All Atoms - Acid Groups " + enumAminoAcidGroups), });
spreadsheet7.Add(AminoAcidChainComposition.SpreadsheetGroupsTitleRow(enumAminoAcidGroups));
var range7 = spreadsheetTaskResult.WholeProteinChainsAminoAcidCounter.Select(record => record.SpreadsheetGroupsDataRow(enumAminoAcidGroups)).ToList();
//range7.Sort();
range7 = range7
.OrderBy(a => a[0].CellData)
.ThenBy(a => a[1].CellData)
.ToList();
spreadsheet7.AddRange(range7);
range7 = null;
spreadsheet7.Add(wholeProteinChainsTotals.SpreadsheetGroupsDataRow(enumAminoAcidGroups));
results.Add(spreadsheet7);
spreadsheet7 = null;
}
}
// convert to percentage for creating mean average protein composition
var meanProteinComposition = new AminoAcidChainComposition("Mean Composition", "-");
foreach (AminoAcidChainComposition aminoAcidChainComposition in spreadsheetTaskResult.WholeProteinChainsAminoAcidCounter)
{
// get percentage for row
AminoAcidChainComposition percentage = AminoAcidChainComposition.ConvertToPercentage(aminoAcidChainComposition);
// add percentage to overall tally
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
for (int x = 0; x < AminoAcidGroups.AminoAcidGroups.GetTotalSubgroups(enumAminoAcidGroups); x++)
{
meanProteinComposition.AminoAcidGroupsCount[(int)enumAminoAcidGroups][x] += (percentage.AminoAcidGroupsCount[(int)enumAminoAcidGroups][x] / spreadsheetTaskResult.WholeProteinChainsAminoAcidCounter.Count);
}
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
{
/* start test */
var spreadsheet8 = new List <SpreadsheetCell[]>();
spreadsheet8.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% TEST SHEET 1"), }); // Worksheet name.
spreadsheet8.Add(new[] { new SpreadsheetCell("TEST SHEET 1"), }); // Spreadsheet title
spreadsheet8.Add(new[] { new SpreadsheetCell(string.Empty), });
spreadsheet8.Add(meanProteinComposition.SpreadsheetDataRow());
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
spreadsheet8.Add(meanProteinComposition.SpreadsheetGroupsDataRow(enumAminoAcidGroups));
}
results.Add(spreadsheet8);
spreadsheet8 = null;
/* end test */
}
AminoAcidPairCompositionMatrix meanProteinMatrix = AminoAcidChainComposition.ConvertToMatrix(meanProteinComposition);
{
var spreadsheet9 = new List <SpreadsheetCell[]>();
spreadsheet9.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% HM All Atoms 3x3"), }); // Worksheet name.
spreadsheet9.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - All Atoms - Average Chain Composition"), }); // Spreadsheet title.
//spreadsheet9.Add(new[] { new SpreadsheetCell(string.Empty), });
//spreadsheet9.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - All Atoms - Average Chain Composition - Percentage Composition - A to Z"), }); // Section title.
//spreadsheet9.AddRange(meanProteinMatrix.SpreadsheetAminoAcidColorGroupsHeatMap());
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
spreadsheet9.Add(new[] { new SpreadsheetCell(string.Empty), });
spreadsheet9.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - All Atoms - Average Chain Composition - Percentage Composition - Acid Groups " + enumAminoAcidGroups), }); // Section title.
spreadsheet9.AddRange(meanProteinMatrix.SpreadsheetAminoAcidColorGroupsHeatMap(enumAminoAcidGroups));
}
results.Add(spreadsheet9);
spreadsheet9 = null;
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//if (outputAllAtoms1x1)
//{
AminoAcidPairCompositionMatrix wholeProteinAminoAcidToAminoAcidCounterPercentage1X1 = AminoAcidPairCompositionMatrix.CalculatePercentageMatrix(wholeProteinAminoAcidToAminoAcidCounter1X1);
{
var spreadsheet10 = new List <SpreadsheetCell[]>();
spreadsheet10.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% HM All Atoms 1x1") }); // Worksheet name.
spreadsheet10.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - All Atoms - Overall Composition") }); // Spreadsheet title.
//spreadsheet10.Add(new[] { new SpreadsheetCell(string.Empty)});
//spreadsheet10.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - All Atoms - Overall Percentage Composition - A to Z")}); // Section title.
//spreadsheet10.AddRange(wholeProteinAminoAcidToAminoAcidCounterPercentage1X1.SpreadsheetAminoAcidColorGroupsHeatMap());
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
spreadsheet10.Add(new[] { new SpreadsheetCell(string.Empty) });
spreadsheet10.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - All Atoms - Overall Percentage Composition - Acid Groups " + enumAminoAcidGroups) }); // Section title.
spreadsheet10.AddRange(wholeProteinAminoAcidToAminoAcidCounterPercentage1X1.SpreadsheetAminoAcidColorGroupsHeatMap(enumAminoAcidGroups));
}
AminoAcidPairCompositionMatrix wholeProteinAminoAcidToAminoAcidCounterNormalised1X1 = AminoAcidPairCompositionMatrix.NormalizeWithCompositionMatrix(wholeProteinAminoAcidToAminoAcidCounterPercentage1X1, UniProtProteinDatabaseComposition.AminoAcidCompositionAsMatrix());
//spreadsheet10.Add(new[] { new SpreadsheetCell(string.Empty)});
//spreadsheet10.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - All Atoms - Overall UniProt Normalised - A to Z ")}); // Section title.
//spreadsheet10.AddRange(wholeProteinAminoAcidToAminoAcidCounterNormalised1X1.SpreadsheetAminoAcidColorGroupsHeatMap());
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
spreadsheet10.Add(new[] { new SpreadsheetCell(string.Empty) });
spreadsheet10.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - All Atoms - Overall UniProt Normalised - Acid Groups " + enumAminoAcidGroups) }); // Section title.
spreadsheet10.AddRange(wholeProteinAminoAcidToAminoAcidCounterNormalised1X1.SpreadsheetAminoAcidColorGroupsHeatMap(enumAminoAcidGroups));
}
AminoAcidPairCompositionMatrix wholeProteinAminoAcidToAminoAcidCounterDifference1X1 = AminoAcidPairCompositionMatrix.DifferenceWithCompositionMatrix(wholeProteinAminoAcidToAminoAcidCounterPercentage1X1, UniProtProteinDatabaseComposition.AminoAcidCompositionAsMatrix());
//spreadsheet10.Add(new[] { new SpreadsheetCell(string.Empty)});
//spreadsheet10.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - All Atoms - Overall A to Z - UniProt Difference")}); // Section title.
//spreadsheet10.AddRange(wholeProteinAminoAcidToAminoAcidCounterDifference1X1.SpreadsheetAminoAcidColorGroupsHeatMap());
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
spreadsheet10.Add(new[] { new SpreadsheetCell(string.Empty) });
spreadsheet10.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - All Atoms - Overall Acid Groups " + enumAminoAcidGroups + " - UniProt Difference") }); // Section title.
spreadsheet10.AddRange(wholeProteinAminoAcidToAminoAcidCounterDifference1X1.SpreadsheetAminoAcidColorGroupsHeatMap(enumAminoAcidGroups));
}
results.Add(spreadsheet10);
spreadsheet10 = null;
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
{
AminoAcidPairCompositionMatrix interactionsAminoAcidToAminoAcidCounterPercentage = AminoAcidPairCompositionMatrix.CalculatePercentageMatrix(spreadsheetTaskResult.InteractionsAminoAcidToAminoAcidCounter);
var spreadsheet11 = new List <SpreadsheetCell[]>();
spreadsheet11.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% HM Interactions Only") }); // Worksheet name.
spreadsheet11.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - Interactions Only") }); // Spreadsheet title.
//spreadsheet11.Add(new[] { new SpreadsheetCell(string.Empty)});
//spreadsheet11.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - Interactions Only - A to Z")}); // Section title.
//spreadsheet11.AddRange(spreadsheetTaskResult.InteractionsAminoAcidToAminoAcidCounter.SpreadsheetAminoAcidColorGroupsHeatMap());
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
spreadsheet11.Add(new[] { new SpreadsheetCell(string.Empty) });
spreadsheet11.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - Interactions Only - Acid Groups " + enumAminoAcidGroups) }); // Section title.
spreadsheet11.AddRange(spreadsheetTaskResult.InteractionsAminoAcidToAminoAcidCounter.SpreadsheetAminoAcidColorGroupsHeatMap(enumAminoAcidGroups));
}
AminoAcidPairCompositionMatrix interactionsAminoAcidToAminoAcidCounterNormalised = AminoAcidPairCompositionMatrix.NormalizeWithCompositionMatrix(interactionsAminoAcidToAminoAcidCounterPercentage, UniProtProteinDatabaseComposition.AminoAcidCompositionAsMatrix());
//spreadsheet11.Add(new[] { new SpreadsheetCell(string.Empty)});
//spreadsheet11.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - Interactions Only - A to Z - UniProt Normalised")}); // Section title.
//spreadsheet11.AddRange(interactionsAminoAcidToAminoAcidCounterNormalised.SpreadsheetAminoAcidColorGroupsHeatMap());
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
spreadsheet11.Add(new[] { new SpreadsheetCell(string.Empty) });
spreadsheet11.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - Interactions Only - Acid Groups " + enumAminoAcidGroups + " - UniProt Normalised") }); // Section title.
spreadsheet11.AddRange(interactionsAminoAcidToAminoAcidCounterNormalised.SpreadsheetAminoAcidColorGroupsHeatMap(enumAminoAcidGroups));
}
AminoAcidPairCompositionMatrix interactionsAminoAcidToAminoAcidCounterDifference = AminoAcidPairCompositionMatrix.DifferenceWithCompositionMatrix(interactionsAminoAcidToAminoAcidCounterPercentage, UniProtProteinDatabaseComposition.AminoAcidCompositionAsMatrix());
//spreadsheet11.Add(new[] { new SpreadsheetCell(string.Empty)});
//spreadsheet11.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - Interactions Only - A to Z - UniProt Difference")}); // Section title.
//spreadsheet11.AddRange(interactionsAminoAcidToAminoAcidCounterDifference.SpreadsheetAminoAcidColorGroupsHeatMap());
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
spreadsheet11.Add(new[] { new SpreadsheetCell(string.Empty) });
spreadsheet11.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - Interactions Only - Acid Groups " + enumAminoAcidGroups + " - UniProt Difference") }); // Section title.
spreadsheet11.AddRange(interactionsAminoAcidToAminoAcidCounterDifference.SpreadsheetAminoAcidColorGroupsHeatMap(enumAminoAcidGroups));
}
results.Add(spreadsheet11);
spreadsheet11 = null;
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
{
var spreadsheet12 = new List <SpreadsheetCell[]>();
spreadsheet12.Add(new[] { new SpreadsheetCell("%batch_number%%batch_letter% HM Interactions v Homodimers") }); // Worksheet name.
spreadsheet12.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - Difference between homodimer composition and homodimer interactions") }); // Spreadsheet title
spreadsheet12.Add(new[] { new SpreadsheetCell(string.Empty) });
//spreadsheet12.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - Difference between homodimer composition and homodimer interactions - A to Z")}); // Section title
//spreadsheet12.AddRange(AminoAcidPairCompositionMatrix.DifferenceWithCompositionMatrix(wholeProteinAminoAcidToAminoAcidCounterPercentage1X1, spreadsheetTaskResult.InteractionsAminoAcidToAminoAcidCounter).SpreadsheetAminoAcidColorGroupsHeatMap());
//spreadsheet12.Add(new[] { new SpreadsheetCell(string.Empty)});
foreach (AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups enumAminoAcidGroups in Enum.GetValues(typeof(AminoAcidGroups.AminoAcidGroups.EnumAminoAcidGroups)))
{
spreadsheet12.Add(new[] { new SpreadsheetCell("Homodimers - Amino Acid Heat Map - Difference between homodimer composition and homodimer interactions - Acid Groups " + enumAminoAcidGroups) }); // Section title.
spreadsheet12.AddRange(AminoAcidPairCompositionMatrix.DifferenceWithCompositionMatrix(wholeProteinAminoAcidToAminoAcidCounterPercentage1X1, spreadsheetTaskResult.InteractionsAminoAcidToAminoAcidCounter).SpreadsheetAminoAcidColorGroupsHeatMap(enumAminoAcidGroups));
spreadsheet12.Add(new[] { new SpreadsheetCell(string.Empty) });
}
results.Add(spreadsheet12);
}
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
return(results);
}
static void Main(string[] args)
{
// this program takes a fasta or pdb file and finds all matching homologs
// FindHomologs.exe "c:\ds96ub\ds96ub.fasta" * "c:\pdb\pdb_seqres.fasta" NMW Y 0.3 75 c:\pdb\
// alignment_type = (n)one, (s)imple, NMW, SWM
var query_sequence_file = args[0]; //query.fasta
var query_id_chain = args[1]; //1A2G:B
var target_sequence_file = args[2]; //targets.fasta
var alignment_type_str = args[3]; //NMW,SWM,SIM,NON
if (alignment_type_str == "*")
{
alignment_type_str = "NMW,SWM,SIM,NON";
}
var alignment_type_str_split = alignment_type_str.ToUpperInvariant().Split(new char[] { ',', ';', ' ', '\t' });
var compare_physicochemically = args[4]; //Y/N
var compare_physicochemically_bool = compare_physicochemically == "Y";
var min_similarity_str = args[5]; // 0.3
var max_len_difference = args[6];
var max_len_difference_int = int.Parse(max_len_difference);
var output_folder = args[7];
var minSimilarity = decimal.Parse(min_similarity_str);
var alignmentTypes = new List <ProteinBioClass.AlignmentType>();
if (alignment_type_str_split.Contains("NMW"))
{
alignmentTypes.Add(ProteinBioClass.AlignmentType.NMW);
}
if (alignment_type_str_split.Contains("SWM"))
{
alignmentTypes.Add(ProteinBioClass.AlignmentType.SWM);
}
if (alignment_type_str_split.Contains("SIM"))
{
alignmentTypes.Add(ProteinBioClass.AlignmentType.SIM);
}
if (alignment_type_str_split.Contains("NON") || alignmentTypes.Count == 0)
{
alignmentTypes.Add(ProteinBioClass.AlignmentType.NON);
}
if (alignmentTypes.Count < alignment_type_str_split.Length)
{
Console.WriteLine("; unknown alignment type");
return;
}
// load list of query sequences
var queryPdbid = query_id_chain.Split(new char[] { ':' })[0];
var queryChainid = (query_id_chain.Contains(":") ? query_id_chain.Split(new char[] { ':' })[1] : "*")[0];
var querySeq = Sequence.LoadSequenceFile(query_sequence_file, null);
var queryResults = querySeq.Where(a =>
{
var id = new ProteinBioClass.SequenceId(a.Id);
return((queryPdbid == "*" || id.PdbId.ToUpperInvariant() == queryPdbid.ToUpperInvariant()) &&
(queryChainid == '*' || id.ChainId == queryChainid));
}).ToList();
if (queryResults.Count == 0)
{
Console.WriteLine("; the query pdbids/chainids were not found");
return;
}
// load list of target sequences
var targetSeq = Sequence.LoadSequenceFile(target_sequence_file, new string[] { null, "", "protein" });
targetSeq = targetSeq.Where(a => a.Count() >= 50).ToList();
Console.WriteLine("; aligning " + queryResults.Count + " query sequences to " + targetSeq.Count + " target sequences");
// perform alignment
//var startTime = DateTime.Now;
//var progress = 0;
//var progressLock = new object();
//var tasks = new List<Task<StringBuilder>>();
var queryPdbIds = queryResults.Select(a => new ProteinBioClass.SequenceId(a.Id).PdbId);
var targetPdbIds = targetSeq.Select(a => new ProteinBioClass.SequenceId(a.Id).PdbId);
var queryPdbIdCounts = new Dictionary <string, int>();
foreach (var x in queryPdbIds)
{
if (!queryPdbIdCounts.ContainsKey(x))
{
queryPdbIdCounts.Add(x, 1);
}
else
{
queryPdbIdCounts[x]++;
}
}
var targetPdbIdCounts = new Dictionary <string, int>();
foreach (var x in targetPdbIds)
{
if (!targetPdbIdCounts.ContainsKey(x))
{
targetPdbIdCounts.Add(x, 1);
}
else
{
targetPdbIdCounts[x]++;
}
}
foreach (var _query in queryResults)
{
var _queryId = new ProteinBioClass.SequenceId(_query.Id);
var filename = (new DirectoryInfo(output_folder).FullName) + @"\homologs_" + _queryId.PdbId + _queryId.ChainId + @".csv";
// skip if already processed
if (File.Exists(filename) && new FileInfo(filename).Length > 0)
{
continue;
}
var totalQueryPdbIdChains = queryPdbIdCounts[_queryId.PdbId];
WorkDivision wd = new WorkDivision(targetSeq.Count);
for (var thread = 0; thread < wd.ThreadCount; thread++)
{
var query = _query;
var queryId = _queryId;
var lti = thread;
wd.TaskList.Add(Task.Run(() =>
{
var result = new List <HomologChain>();
for (var target = wd.ThreadFirstIndex[lti]; target <= wd.ThreadLastIndex[lti]; target++)
{
var targetobj = targetSeq[target];
if (max_len_difference_int != -1 && Math.Abs(targetobj.Count() - query.Count()) > max_len_difference_int)
{
continue;
}
var targetId = new ProteinBioClass.SequenceId(targetobj.Id);
//var totalTargetPdbIdChains = targetSeq.Count(a => FindAtomicContacts.SequenceIdToPdbIdAndChainId(a.ID).PdbId.ToUpperInvariant() == targetId.PdbId.ToUpperInvariant());
//var timeRemaining =
// TimeSpan.FromTicks(DateTime.Now.Subtract(startTime).Ticks *
// ((targetSeq.Count * queryResults.Count) - (progress + 1)) /
// (progress + 1));
foreach (var alignmentType in alignmentTypes)
{
var scores = ProteinBioClass.AlignedSequenceSimilarityPercentage(query, targetobj, alignmentType /*,
* compare_physicochemically_bool*/);
decimal percentSimilar;
if (compare_physicochemically_bool)
{
percentSimilar = scores.ScoreEvo;
}
else
{
percentSimilar = scores.Score;
}
if (percentSimilar >= minSimilarity)
{
result.Add(new HomologChain(
queryId.PdbId, queryId.ChainId, totalQueryPdbIdChains,
targetId.PdbId, targetId.ChainId, targetPdbIdCounts[targetId.PdbId],
alignmentType.ToString(),
scores.Score,
scores.ScoreEvo));
}
}
//if (progress % 20 == 0)
// Console.Write("\r{0}% eta {1} ",
// Math.Round((decimal)(progress + 1) / (decimal)(targetSeq.Count * queryResults.Count),
// 2)
// .ToString(CultureInfo.InvariantCulture),
// timeRemaining.ToString(@"d\d\:h\h\:m\m\:s\s",
// CultureInfo.InvariantCulture));
//lock (progressLock)
// progress++;
}
return(result);
}));
}
wd.WaitAllTasks();
var mergedlist = new List <string>();
mergedlist.Add("; " + _queryId.PdbId + ":" + _queryId.ChainId);
mergedlist.Add(String.Join(",",
new string[]
{
"query pdb id", "query chain id", "query chains",
"target pdb id", "target chain id", "target chains",
"alignment method", "sequence similarity", "sequence evo similarity"
}));
foreach (var t in wd.TaskList)
{
var tc = t as Task <List <HomologChain> >;
if (tc == null)
{
throw new Exception("task in tasklist was null");
}
mergedlist.AddRange(tc.Result.Select(a => a.ToString()).ToList());
}
if (string.IsNullOrWhiteSpace(output_folder))
{
Console.WriteLine(String.Join(Environment.NewLine, mergedlist));
}
else
{
File.WriteAllLines(filename, mergedlist);
}
}
}
public static List <AtomPair> FindInteractions(CancellationToken cancellationToken, decimal maxAtomInterationDistance /*= 8.0m*/, string proteinId, Dictionary <string, List <string> > pdbIdChainIdList, ProteinChainListContainer proteinFileChains, bool breakWhenFirstInteractionFound = false, int totalThreads = -1, bool sort = true, int requiredChains = -1)
{
//const decimal maxInterationDistance = 8.0m;
bool useCache = false;
if (useCache && !string.IsNullOrWhiteSpace(proteinId))
{
var cachedInteractions = InteractionsCache.LoadPdbInteractionCache(proteinId, requiredChains);
if (cachedInteractions != null)
{
return(cachedInteractions);
}
}
// check required number of chains are found
if (proteinFileChains == null || proteinFileChains.ChainList == null || (requiredChains > -1 && proteinFileChains.ChainList.Count != requiredChains))
{
return(null);
}
// check that all chains have atoms
if (proteinFileChains.ChainList.Any(chain => chain.AtomList == null || chain.AtomList.Count == 0))
{
return(null);
}
// Make list of 3D positions of atoms.
var positions = new List <Point3D> [proteinFileChains.ChainList.Count];
for (int chainIndex = 0; chainIndex < proteinFileChains.ChainList.Count; chainIndex++)
{
positions[chainIndex] = Clustering.AtomRecordListToPoint3DList(proteinFileChains.ChainList[chainIndex]);
}
var tasks = new List <Task <List <AtomPair> > >();
for (int chainIndexA = 0; chainIndexA < proteinFileChains.ChainList.Count; chainIndexA++)
{
for (int chainIndexB = 0; chainIndexB < proteinFileChains.ChainList.Count; chainIndexB++)
{
if (chainIndexB == chainIndexA || chainIndexB < chainIndexA)
{
continue;
}
WorkDivision <List <AtomPair> > workDivision = new WorkDivision <List <AtomPair> >(proteinFileChains.ChainList[chainIndexA].AtomList.Count, totalThreads);
bool breakOut = false;
var lockBreakOut = new object();
for (int threadIndex = 0; threadIndex < workDivision.ThreadCount; threadIndex++)
{
int localThreadIndex = threadIndex;
int localChainIndexA = chainIndexA;
int localChainIndexB = chainIndexB;
WorkDivision <List <AtomPair> > localWorkDivision = workDivision;
Task <List <AtomPair> > task = Task.Run(() =>
{
var taskResult = new List <AtomPair>();
for (int atomIndexA = localWorkDivision.ThreadFirstIndex[localThreadIndex]; atomIndexA <= localWorkDivision.ThreadLastIndex[localThreadIndex]; atomIndexA++)
{
if (breakOut)
{
break;
}
for (int atomIndexB = 0; atomIndexB < proteinFileChains.ChainList[localChainIndexB].AtomList.Count; atomIndexB++)
{
if (breakOut || (breakWhenFirstInteractionFound && taskResult.Count > 0))
{
lock (lockBreakOut)
{
breakOut = true;
}
break;
}
if ((!positions[localChainIndexA][atomIndexA].ParseOK) || (!positions[localChainIndexB][atomIndexB].ParseOK))
{
continue;
}
decimal atomicDistanceAngstroms3D = Point3D.Distance3D(positions[localChainIndexA][atomIndexA], positions[localChainIndexB][atomIndexB], true);
// Chemical proteinInterface bonds found at 5 angstrom or less.
if (atomicDistanceAngstroms3D <= 0.0m || atomicDistanceAngstroms3D > maxAtomInterationDistance)
{
continue;
}
var atomPair = new AtomPair(
proteinId,
proteinFileChains.ChainList[localChainIndexA].AtomList[atomIndexA],
localChainIndexA,
proteinId,
localChainIndexB,
proteinFileChains.ChainList[localChainIndexB].AtomList[atomIndexB],
atomicDistanceAngstroms3D);
taskResult.Add(atomPair);
}
}
if (taskResult.Count == 0)
{
return(null);
}
return(taskResult);
}, cancellationToken);
workDivision.TaskList.Add(task);
}
tasks.AddRange(workDivision.TaskList);
}
}
try
{
Task[] tasksToWait = tasks.Where(task => task != null && !task.IsCompleted).ToArray <Task>();
if (tasksToWait.Length > 0)
{
Task.WaitAll(tasksToWait);
}
}
catch (AggregateException)
{
}
// merge all results
var atomPairList = new List <AtomPair>();
foreach (var task in tasks.Where(t => t != null && t.IsCompleted && !t.IsCanceled && !t.IsFaulted && t.Result != null && t.Result.Count > 0))
{
atomPairList.AddRange(task.Result);
}
if (sort && atomPairList != null && atomPairList.Count > 1)
{
atomPairList = atomPairList
.OrderBy(i => ProteinDataBankFileOperations.NullableTryParseInt32(i.Atom1.resSeq.FieldValue))
.ThenBy(i => ProteinDataBankFileOperations.NullableTryParseInt32(i.Atom1.serial.FieldValue))
.ThenBy(j => ProteinDataBankFileOperations.NullableTryParseInt32(j.Atom2.resSeq.FieldValue))
.ThenBy(j => ProteinDataBankFileOperations.NullableTryParseInt32(j.Atom2.serial.FieldValue))
.ToList();
}
if (useCache)
{
InteractionsCache.SavePdbInteractionCache(proteinId, atomPairList, requiredChains);
}
return(atomPairList);
}
/*
* public static void ClusterVectorDistanceMatrixUpgma(List<VectorProteinInterfaceWhole> vectorProteinInterfaceWholeList, decimal[,] vectorDistanceMatrix, int minimumOutputTreeLeafs, out List<string> vectorNames, out List<List<UpgmaNode>> nodeList, out List<List<string>> treeList, ProgressActionSet progressActionSet)
* {
* if (vectorProteinInterfaceWholeList == null) throw new ArgumentNullException(nameof(vectorProteinInterfaceWholeList));
* if (vectorDistanceMatrix == null) throw new ArgumentNullException(nameof(vectorDistanceMatrix));
*
* vectorNames = vectorProteinInterfaceWholeList.Select(VectorProteinInterfaceWholeTreeHeader).ToList();
*
* List<string> finalTreeLeafOrderList;
* UpgmaClustering.Upgma(vectorDistanceMatrix, vectorNames, minimumOutputTreeLeafs, out nodeList, out treeList, out finalTreeLeafOrderList, false, progressActionSet);
* }
*/
public static void BestDistanceMatrixWithPartsAlignment(CancellationToken cancellationToken, List <VectorProteinInterfaceWhole> vectorProteinInterfaceWholeList, VectorDistanceMeasurementValues vectorDistanceMeasurementValues, out double[,] optimisticDistanceMatrix, /* out double[,] pessimisticDistanceMatrix,*/ ProgressActionSet progressActionSet)
{
if (vectorProteinInterfaceWholeList == null)
{
throw new ArgumentNullException(nameof(vectorProteinInterfaceWholeList));
}
if (vectorDistanceMeasurementValues == null)
{
throw new ArgumentNullException(nameof(vectorDistanceMeasurementValues));
}
var totalVectors = vectorProteinInterfaceWholeList.Count;
var optimisticDistanceMatrix2 = new double[totalVectors, totalVectors];
//var pessimisticDistanceMatrix2 = new double[totalVectors, totalVectors];
var workDivision = new WorkDivision(vectorProteinInterfaceWholeList.Count, -1);
ProgressActionSet.StartAction(vectorProteinInterfaceWholeList.Count, progressActionSet);
for (int threadIndex = 0; threadIndex < workDivision.ThreadCount; threadIndex++)
{
int localThreadIndex = threadIndex;
var task = Task.Run(() =>
{
for (int indexX = workDivision.ThreadFirstIndex[localThreadIndex]; indexX <= workDivision.ThreadLastIndex[localThreadIndex]; indexX++)
{
if (cancellationToken.IsCancellationRequested)
{
break;
}
var vectorProteinInterfaceWholeX = vectorProteinInterfaceWholeList[indexX];
for (int indexY = 0; indexY < vectorProteinInterfaceWholeList.Count; indexY++)
{
if (indexX >= indexY)
{
continue;
}
var vectorProteinInterfaceWholeY = vectorProteinInterfaceWholeList[indexY];
if (vectorProteinInterfaceWholeX.FullProteinInterfaceId == vectorProteinInterfaceWholeY.FullProteinInterfaceId)
{
continue;
}
double optimisticDistance;
//double pessimisticDistance;
BestDistanceWithPartsAlignment(vectorProteinInterfaceWholeX, vectorProteinInterfaceWholeY, vectorDistanceMeasurementValues, out optimisticDistance /*, out pessimisticDistance*/);
var lengthDifference = Math.Abs(vectorProteinInterfaceWholeX.ProteinInterfaceLength - vectorProteinInterfaceWholeY.ProteinInterfaceLength);
var lengthDistance = lengthDifference * vectorDistanceMeasurementValues.DifferentLengthProteinInterface;
optimisticDistance += lengthDistance;
//pessimisticDistance += lengthDistance;
optimisticDistanceMatrix2[indexX, indexY] = optimisticDistance;
//pessimisticDistanceMatrix2[indexX, indexY] = pessimisticDistance;
optimisticDistanceMatrix2[indexY, indexX] = optimisticDistance;
//pessimisticDistanceMatrix2[indexY, indexX] = pessimisticDistance;
}
workDivision.IncrementItemsCompleted(1);
ProgressActionSet.ProgressAction(1, progressActionSet);
ProgressActionSet.EstimatedTimeRemainingAction(workDivision.StartTicks, workDivision.ItemsCompleted, workDivision.ItemsToProcess, progressActionSet);
}
}, cancellationToken);
workDivision.TaskList.Add(task);
}
workDivision.WaitAllTasks();
ProgressActionSet.FinishAction(true, progressActionSet);
optimisticDistanceMatrix = optimisticDistanceMatrix2;
//pessimisticDistanceMatrix = pessimisticDistanceMatrix2;
}
private static void Main(string[] args)
{
// this program will load the homolog list in csv format and for homologs of X sequence distance return a list of all partners
// however, some partners may be duplicates, which cannot initially be removed, since they may bind differently in other instances
// then, because of such cases, unique id to describe each protein must be created... this is slightly problematic because
// close target homologs of proteins are also considered to be the same protein as the query protein
// which means that they could exist for more than one query protein
// FindHomologsCluster.exe c:\pdb\ds96ub_homologs\ c:\pdb\pdb_seqres.fasta 0.9 0.9 > ds96ub_homologs.csv
var homolog_csv_folder = args[0];
var sequence_file = args[1];
var min_similarity_str = args[2];
var min_similarity_evo_str = args[3];
var min_similarity = decimal.Parse(min_similarity_str);
var min_similarity_evo = decimal.Parse(min_similarity_evo_str);
var seqList = Sequence.LoadSequenceFile(sequence_file, new[] { null, "", "protein" });
var homologCsvFiles = Directory.GetFiles(homolog_csv_folder, "homologs_?????.csv");
var parsedData = FindHomologs.FindHomologs.HomologChain.Load(homologCsvFiles);
Array.Clear(homologCsvFiles, 0, homologCsvFiles.Length);
//var query_pdb_list = parsed_data.Select(a => new Tuple<string, string>(a.query_pdb_id, a.query_chainid)).ToList();
//var target_pdb_list = parsed_data.Select(a => new Tuple<string, string>(a.target_pdb_id, a.target_chainid)).ToList();
//var query_alignments = new List<homolog_csv>();
var homologs_clustered = new List <List <Tuple <string, char> > >();
//var min_similarity = 0.9m;
foreach (var rec in parsedData)
{
if (rec.AlignmentScore >= min_similarity && rec.AlignmentScoreEvo >= min_similarity_evo)
{
//var query_group = homologs_clustered.FirstOrDefault(a => a.FirstOrDefault(b => b.Item1 == rec.query_pdb_id && b.Item2 == rec.query_chainid) != null);
//var target_group = homologs_clustered.FirstOrDefault(a => a.FirstOrDefault(b => b.Item1 == rec.target_pdb_id && b.Item2 == rec.target_chainid) != null);
List <Tuple <string, char> > query_group = null;
List <Tuple <string, char> > target_group = null;
foreach (var cluster in homologs_clustered)
{
var xq = cluster.FirstOrDefault(b => b.Item1.ToUpperInvariant() == rec.QueryPdbId.ToUpperInvariant() && b.Item2 == rec.QueryChainId);
if (xq == null)
{
continue;
}
query_group = cluster;
break;
}
foreach (var cluster in homologs_clustered)
{
var xt =
cluster.FirstOrDefault(b => b.Item1.ToUpperInvariant() == rec.TargetPdbId.ToUpperInvariant() && b.Item2 == rec.TargetChainId);
if (xt == null)
{
continue;
}
target_group = cluster;
break;
}
var new_group = new List <Tuple <string, char> >();
if (query_group != null)
{
new_group.AddRange(query_group);
homologs_clustered.Remove(query_group);
query_group.Clear();
}
else
{
new_group.Add(new Tuple <string, char>(rec.QueryPdbId, rec.QueryChainId));
}
if (target_group != null)
{
new_group.AddRange(target_group);
homologs_clustered.Remove(target_group);
target_group.Clear();
}
else
{
new_group.Add(new Tuple <string, char>(rec.TargetPdbId, rec.TargetChainId));
}
new_group = new_group.Distinct().ToList(); // try without distinct?
new_group = new_group.OrderBy(a => a.Item1).ThenBy(a => a.Item2).ToList();
homologs_clustered.Add(new_group);
}
}
var seq_list_ids = seqList.Select(a => new ProteinBioClass.SequenceId(a.Id)).ToList();
var wd2 = new WorkDivision(homologs_clustered.Count);
for (var thread2 = 0; thread2 < wd2.ThreadCount; thread2++)
{
var lti2 = thread2;
wd2.TaskList.Add(Task.Run(() =>
{
var result2 = new List <string>();
for (var index2 = wd2.ThreadFirstIndex[lti2]; index2 <= wd2.ThreadLastIndex[lti2]; index2++)
{
var cluster2 = homologs_clustered[index2];
var wd3 = new WorkDivision(cluster2.Count);
for (var thread3 = 0; thread3 < wd3.ThreadCount; thread3++)
{
var lti3 = thread3;
var cluster3 = cluster2;
var index4 = index2;
wd3.TaskList.Add(Task.Run(() =>
{
var result = new List <HomologClusterData>();
for (var index3 = wd3.ThreadFirstIndex[lti3]; index3 <= wd3.ThreadLastIndex[lti3]; index3++)
{
var item = cluster3[index3];
Sequence s = null;
for (var j = 0; j < seqList.Count; j++)
{
if (seq_list_ids[j].PdbId.ToUpperInvariant() == item.Item1.ToUpperInvariant() && seq_list_ids[j].ChainId == item.Item2)
{
s = seqList[j];
break;
}
}
if (s == null)
{
throw new Exception("sequence not found for " + item.Item1 + ":" + item.Item2);
}
var complexChains = seq_list_ids.Count(a => a.PdbId.ToUpperInvariant() == item.Item1.ToUpperInvariant());
var minAlignmentScore = -1m;
var maxAlignmentScore = -1m;
var minAlignmentScoreEvo = -1m;
var maxAlignmentScoreEvo = -1m;
foreach (var item2 in cluster3)
{
if (ReferenceEquals(item, item2))
{
continue;
}
Sequence s2 = null;
for (var j2 = 0; j2 < seqList.Count; j2++)
{
if (seq_list_ids[j2].PdbId.ToUpperInvariant() == item2.Item1.ToUpperInvariant() &&
seq_list_ids[j2].ChainId == item2.Item2)
{
s2 = seqList[j2];
break;
}
}
if (s2 == null)
{
continue;
}
var alignmentScore = ProteinBioClass.AlignedSequenceSimilarityPercentage(s,
s2,
ProteinBioClass.AlignmentType.NMW);
if (alignmentScore.Score > maxAlignmentScore || maxAlignmentScore == -1m)
{
maxAlignmentScore = alignmentScore.Score;
}
if (alignmentScore.Score < minAlignmentScore || minAlignmentScore == -1m)
{
minAlignmentScore = alignmentScore.Score;
}
if (alignmentScore.ScoreEvo > maxAlignmentScoreEvo || maxAlignmentScoreEvo == -1m)
{
maxAlignmentScoreEvo = alignmentScore.ScoreEvo;
}
if (alignmentScore.ScoreEvo < minAlignmentScoreEvo || minAlignmentScoreEvo == -1m)
{
minAlignmentScoreEvo = alignmentScore.ScoreEvo;
}
}
var r = new HomologClusterData(index4 + 1, index3 + 1, item.Item1, item.Item2, complexChains, Convert.ToInt32(s.Count()), minAlignmentScore, maxAlignmentScore, minAlignmentScoreEvo, maxAlignmentScoreEvo, s.FullSequence);
result.Add(r);
}
return(result);
}));
}
wd3.WaitAllTasks();
result2.Add("; Cluster # " + (index2 + 1) + " with " + wd3.ItemsToProcess + " protein chains");
result2.Add("cluster index,item index,pdb id,chain id,complex chains,seq len,min clstr sid,max clstr sid,min evo clstr sid,max evo clstr sid,sequence");
foreach (var task in wd3.TaskList)
{
//if (task.IsFaulted || task.IsCanceled) continue;
var tr = task as Task <List <HomologClusterData> >;
if (tr == null || tr.Result == null)
{
continue;
}
result2.AddRange(tr.Result.Select(a => a.ToString()).ToList());
}
result2.Add("");
}
return(result2);
}));
//wd2.TaskList.Add(task2);
}
wd2.WaitAllTasks();
var result1 = new List <string>();
foreach (var task in wd2.TaskList)
{
//if (task.IsFaulted || task.IsCanceled) continue;
var tr = task as Task <List <string> >;
if (tr == null || tr.Result == null)
{
continue;
}
result1.AddRange(tr.Result);
}
foreach (var line in result1)
{
Console.WriteLine(line);
}
// partners may have other interfaces, should those also be considered?
}
