public AssemblyReport(MitoPaintedAssembler toReportOn, AlgorithmResult result = AlgorithmResult.Success) : base(result)
{
HeaderLineForCSV = CreateHeaderLine();
this.DataLineForCSV = String.Join(",", outputValues.Select(x => x.GetValue(toReportOn).ToString()).ToArray());
}
protected AssemblyReport CreateAssemblyAndDepthOfCoverage()
{
if (Skip_Assembly_Step)
{
return(new AssemblyReport());
}
DepthOfCoverageGraphMaker coveragePlotter = !Skip_DepthOfCoveragePlot ?
new DepthOfCoverageGraphMaker() : null;
IEnumerable <ISequence> reads = this.createSequenceProducer(this.Filename, coveragePlotter, true);
TimeSpan algorithmSpan = new TimeSpan();
Stopwatch runAlgorithm = new Stopwatch();
//Step 1: Initialize assembler.
Output.WriteLine(OutputLevel.Verbose, "\nAssemblying mtDNA and obtaining depth of coverage (if asked).");
MitoPaintedAssembler.StatusChanged += this.StatusChanged;
MitoPaintedAssembler assembler = new MitoPaintedAssembler()
{
DiagnosticFileOutputPrefix = DiagnosticFilePrefix,
AllowErosion = AllowErosion,
AlternateMinimumNodeCount = MinimumNodeCount,
DanglingLinksThreshold = DangleThreshold,
ErosionThreshold = ErosionThreshold,
AllowKmerLengthEstimation = AllowKmerLengthEstimation,
RedundantPathLengthThreshold = RedundantPathLengthThreshold,
OutputIntermediateGraphSteps = OutputIntermediateGraphSteps,
NoContigOutput = NoContigOutput,
ForceSqrtThreshold = ForceSqrtThreshold
};
if (ContigCoverageThreshold != -1)
{
assembler.AllowLowCoverageContigRemoval = true;
assembler.ContigCoverageThreshold = ContigCoverageThreshold;
}
if (!this.AllowKmerLengthEstimation)
{
assembler.KmerLength = this.KmerLength;
}
//Step 2: Assemble
runAlgorithm.Restart();
var assembly = assembler.Assemble(reads);
runAlgorithm.Stop();
algorithmSpan = algorithmSpan.Add(runAlgorithm.Elapsed);
if (this.Verbose)
{
Output.WriteLine(OutputLevel.Verbose);
Output.WriteLine(OutputLevel.Verbose, "\tCompute time: {0}", runAlgorithm.Elapsed);
}
//Step 3: Report
if (!NoContigOutput)
{
runAlgorithm.Restart();
this.writeContigs(assembly);
runAlgorithm.Stop();
}
algorithmSpan = algorithmSpan.Add(runAlgorithm.Elapsed);
if (this.Verbose)
{
Output.WriteLine(OutputLevel.Verbose);
Output.WriteLine(OutputLevel.Verbose, "\tWrite contigs time: {0}", runAlgorithm.Elapsed);
Output.WriteLine(OutputLevel.Verbose, "\tTotal assembly runtime: {0}", algorithmSpan);
}
if (coveragePlotter != null)
{
coveragePlotter.OutputCoverageGraphAndCSV(DiagnosticFilePrefix);
}
return(assembler.GetReport());
}
/// <summary>
/// Assembles the sequences and returns the string that can be placed in a CSV output report.
/// </summary>
/// <returns></returns>
public string AssembleSequencesReturningString()
{
string toReturn = "No String Set";
TimeSpan algorithmSpan = new TimeSpan();
Stopwatch runAlgorithm = new Stopwatch();
FileInfo refFileinfo = new FileInfo(this.Filename);
long refFileLength = refFileinfo.Length;
runAlgorithm.Restart();
IEnumerable <ISequence> reads = this.ParseFile();
runAlgorithm.Stop();
algorithmSpan = algorithmSpan.Add(runAlgorithm.Elapsed);
Output.WriteLine(OutputLevel.Information, StaticResources.AssemblyStarting);
if (this.Verbose)
{
Output.WriteLine(OutputLevel.Verbose);
Output.WriteLine(OutputLevel.Verbose, "Processed read file: {0}", Path.GetFullPath(this.Filename));
Output.WriteLine(OutputLevel.Verbose, " Read/Processing time: {0}", runAlgorithm.Elapsed);
Output.WriteLine(OutputLevel.Verbose, " File Size : {0}", refFileLength);
Output.WriteLine(OutputLevel.Verbose, " k-mer Length : {0}", this.KmerLength);
}
using (MitoPaintedAssembler assembler = new MitoPaintedAssembler())
{
assembler.AllowErosion = true;
//assembler.ReferenceGenomeFile = ReferenceGenome;
assembler.DiagnosticFileOutputPrefix = DiagnosticFilePrefix;
Console.WriteLine("Prefix is: " + assembler.DiagnosticFileOutputPrefix);
Console.WriteLine("Diagnostic Information On: " + assembler.OutputDiagnosticInformation.ToString());
assembler.StatusChanged += this.AssemblerStatusChanged;
assembler.AllowErosion = this.AllowErosion;
assembler.AllowKmerLengthEstimation = this.AllowKmerLengthEstimation;
if (ContigCoverageThreshold != -1)
{
assembler.AllowLowCoverageContigRemoval = true;
assembler.ContigCoverageThreshold = ContigCoverageThreshold;
}
assembler.DanglingLinksThreshold = this.DangleThreshold;
assembler.ErosionThreshold = this.ErosionThreshold;
if (!this.AllowKmerLengthEstimation)
{
assembler.KmerLength = this.KmerLength;
}
assembler.RedundantPathLengthThreshold = this.RedundantPathLengthThreshold;
runAlgorithm.Restart();
IDeNovoAssembly assembly = assembler.Assemble(reads);
runAlgorithm.Stop();
algorithmSpan = algorithmSpan.Add(runAlgorithm.Elapsed);
if (this.Verbose)
{
Output.WriteLine(OutputLevel.Verbose);
Output.WriteLine(OutputLevel.Verbose, "Compute time: {0}", runAlgorithm.Elapsed);
}
runAlgorithm.Restart();
this.WriteContigs(assembly);
runAlgorithm.Stop();
toReturn = assembler.GetReportLine();
if (assembler.OutputDiagnosticInformation)
{
var outFile = new StreamWriter(ReportOutputPrefix + DiagnosticFilePrefix + ".csv");
outFile.WriteLine(MitoDataAssembler.AssemblyOutput.CreateHeaderLine());
outFile.WriteLine(toReturn);
outFile.Close();
}
algorithmSpan = algorithmSpan.Add(runAlgorithm.Elapsed);
if (this.Verbose)
{
Output.WriteLine(OutputLevel.Verbose);
Output.WriteLine(OutputLevel.Verbose, "Write contigs time: {0}", runAlgorithm.Elapsed);
Output.WriteLine(OutputLevel.Verbose, "Total runtime: {0}", algorithmSpan);
}
}
return(toReturn);
}
private void attemptToCreateAssembly()
{
//TODO: This node should always be a good start node, but may be an erroneous one, check for this.
var curNode = gg.MetaNodes.Where(x => x.Lowest_Reference_Position != 0).MaxBy(x => (x.AvgKmerCoverage * x.ConstituentNodes.Count));//*(.2/x.Lowest_Reference_Position));//.MinBy(x => x.Lowest_Reference_Position);
//Let's try just going with the forward primer
//var match = forwardPrimer.Substring(0, gg.MegaNodes.First().LeadingKmer.Length);
//var rc_match = ((new Bio.Sequence(Bio.Alphabets.NoGapDNA, match)).GetReverseComplementedSequence() as Bio.Sequence).ConvertToString();
//var curNode = gg.MegaNodes.Where(x => x.Sequence.Contains(match) || x.Sequence.Contains(rc_match)).First();
_greedyPathAssembly = new PossibleAssembly();
if (!curNode.CircularLoop)
{
MitoPaintedAssembler.RaiseStatusEvent("\tAttempting to find greedy path, frequencies of majority split below");
//now to attempt to loop back to the start node
//will move along while greedily grabbing the next node with the highest kmer coverage
//constantly oriented everyone so we go right ot left
while (true)
{
assemblyNodes.Add(curNode);
_greedyPathAssembly.AddMetaNode(curNode);
var possibles = curNode.GetOutgoingNodes().ToList();
if (possibles.Count > 0)
{
SplitData sd = new SplitData(possibles);
PathSplits.Add(sd);
if (possibles.Count > 1)
{
if (sd.MaxFrequency < MinimumGreedySplit)
{
MinimumGreedySplit = sd.MaxFrequency;
}
MitoPaintedAssembler.RaiseStatusEvent("\tPossible Paths: " + possibles.Count
+ " Frequency: " + sd.MaxFrequency.ToString("P1")
+ " Range: " + curNode.Lowest_Reference_Position.ToString() + "-" + curNode.Highest_Reference_Position.ToString());
}
curNode = sd.BestPath.NeighborNode;
if (assemblyNodes.Contains(curNode))
{
FormsCompleteLoop = true;
break;
}
}
else
{
FormsCompleteLoop = false; SuccessfulAssembly = false; break;
}
}
}
else
{
FormsCompleteLoop = true;
assemblyNodes.Add(curNode);
_greedyPathAssembly.AddMetaNode(curNode);
MinimumGreedySplit = 1.0;
}
int length = assemblyNodes.Sum(x => x.LengthOfNode);
//now, did we form an assembly?
if (FormsCompleteLoop || Math.Abs(length - AssemblyLength) < 100)
{
SuccessfulAssembly = true;
_greedyPathAssembly.FinalizeAndOrientToReference();
AssemblyLength = (int)_greedyPathAssembly.Sequence.Count;
//TODO: More sophisticated criteria than larger than 8 kb to validate assembly
if (AssemblyLength > StaticResources.SIZE_DIF_BETWEEN_LARGE_AND_SMALL_DELETION)
{
SuccessfulAssembly = true;
MitoPaintedAssembler.RaiseStatusEvent("\tSuccessful assembly of length: " + AssemblyLength.ToString());
}
else
{
SuccessfulAssembly = false;
MitoPaintedAssembler.RaiseStatusEvent("\tAssembly failed. Only recovered sequence of length: " + AssemblyLength.ToString());
}
}
}