public static List<FeatureItem> GroupByName(this IEnumerable<FeatureLocation> locations)
{
var result = new List<FeatureItem>();
foreach (var curregions in locations.GroupBy(m => m.Name).ToList())
{
var mi = new FeatureItem();
mi.Name = curregions.Key;
mi.Locations.AddRange(curregions);
result.Add(mi);
}
return result;
}
public static void CombineLocationByMappedReads(this FeatureItem item)
{
//deal with the item with multiple regions but one of them contains others
if (item.Locations.Count > 1)
{
var removed = new List <FeatureLocation>();
for (int i = 0; i < item.Locations.Count; i++)
{
var regi = item.Locations[i];
for (int j = i + 1; j < item.Locations.Count; j++)
{
var regj = item.Locations[j];
if (removed.Contains(regj))
{
continue;
}
var con = regi.Contains(regj);
//Keep the small one if two ranges mapped by same reads
if (con == 1)
{
//if i contains j and all mapped reads mapped to j, remove i
if (regi.SamLocations.All(m => regj.SamLocations.Any(l => l.SamLocation == m.SamLocation)))
{
removed.Add(regi);
break;
}
//if i contains j and all mapped reads from j were contained in i, remove j
if (regj.SamLocations.All(m => regi.SamLocations.Any(l => l.SamLocation == m.SamLocation)))
{
removed.Add(regj);
continue;
}
}
else if (con == -1)
{
//if j contains i and all mapped reads mapped to i, remove j
if (regj.SamLocations.All(m => regi.SamLocations.Any(l => l.SamLocation == m.SamLocation)))
{
removed.Add(regj);
continue;
}
//if j contains i and all mapped reads from i were contained in j, remove i
if (regi.SamLocations.All(m => regj.SamLocations.Any(l => l.SamLocation == m.SamLocation)))
{
removed.Add(regi);
break;
}
}
}
}
foreach (var floc in removed)
{
foreach (var sloc in floc.SamLocations)
{
sloc.SamLocation.Parent.RemoveLocation(sloc.SamLocation);
}
}
item.Locations.RemoveAll(m => removed.Contains(m));
}
}
public virtual List <FeatureItemGroup> ReadFromFile(string fileName)
{
var result = new List <FeatureItemGroup>();
using (XmlReader source = XmlReader.Create(fileName))
{
Progress.SetMessage("reading queries ...");
List <SAMAlignedItem> queries = SAMAlignedItemUtils.ReadFrom(source);
Progress.SetMessage("{0} queries read.", queries.Count);
var qmmap = queries.ToSAMAlignedLocationMap();
queries.Clear();
Progress.SetMessage("reading subjects ...");
string value;
source.ReadToFollowing("subjectResult");
if (source.ReadToDescendant("subjectGroup"))
{
do
{
var featureGroup = new FeatureItemGroup();
result.Add(featureGroup);
if (source.ReadToDescendant("subject"))
{
do
{
var item = new FeatureItem();
featureGroup.Add(item);
item.Name = source.GetAttribute("name");
if (source.ReadToDescendant("region"))
{
do
{
var fl = new FeatureLocation();
item.Locations.Add(fl);
fl.Name = item.Name;
fl.Seqname = source.GetAttribute("seqname");
fl.Start = long.Parse(source.GetAttribute("start"));
fl.End = long.Parse(source.GetAttribute("end"));
fl.Strand = source.GetAttribute("strand")[0];
fl.Sequence = source.GetAttribute("sequence");
value = source.GetAttribute("query_count_before_filter");
if (value != null)
{
fl.QueryCountBeforeFilter = int.Parse(value);
}
value = source.GetAttribute("pvalue");
if (value != null)
{
fl.PValue = double.Parse(value);
}
if (source.ReadToDescendant("query"))
{
do
{
string qname = source.GetAttribute("qname");
string loc = source.GetAttribute("loc");
string key = SAMAlignedLocation.GetKey(qname, loc);
SAMAlignedLocation query = qmmap[key];
FeatureSamLocation fsl = new FeatureSamLocation(fl);
fsl.SamLocation = query;
fsl.Offset = int.Parse(source.GetAttribute("offset"));
var attr = source.GetAttribute("overlap");
if (attr == null)
{
fsl.OverlapPercentage = query.OverlapPercentage(fl);
}
else
{
fsl.OverlapPercentage = double.Parse(attr);
}
var nmi = source.GetAttribute("nmi");
if (nmi != null)
{
fsl.NumberOfMismatch = int.Parse(nmi);
}
var nnpm = source.GetAttribute("nnpm");
if (nnpm != null)
{
fsl.NumberOfNoPenaltyMutation = int.Parse(nnpm);
}
} while (source.ReadToNextSibling("query"));
}
} while (source.ReadToNextSibling("region"));
}
} while (source.ReadToNextSibling("subject"));
}
} while (source.ReadToNextSibling("subjectGroup"));
}
qmmap.Clear();
}
Progress.SetMessage("{0} subjects read.", result.Count);
return(result);
}
public List <FeatureItemGroup> ReadFromFile(string fileName)
{
Console.WriteLine("read file {0} ...", fileName);
var result = new List <FeatureItemGroup>();
XElement root = XElement.Load(fileName);
//Console.WriteLine("read locations ...");
Dictionary <string, SAMAlignedLocation> qmmap = root.ToSAMAlignedItems().ToSAMAlignedLocationMap();
//Console.WriteLine("read mapped items ...");
foreach (XElement groupEle in root.Element("subjectResult").Elements("subjectGroup"))
{
var group = new FeatureItemGroup();
result.Add(group);
foreach (XElement featureEle in groupEle.Elements("subject"))
{
var item = new FeatureItem();
group.Add(item);
item.Name = featureEle.Attribute("name").Value;
foreach (XElement locEle in featureEle.Elements("region"))
{
var fl = new FeatureLocation();
item.Locations.Add(fl);
fl.Name = item.Name;
fl.ParseLocation(locEle);
if (locEle.Attribute("sequence") != null)
{
fl.Sequence = locEle.Attribute("sequence").Value;
}
if (locEle.Attribute("query_count_before_filter") != null)
{
fl.QueryCountBeforeFilter = int.Parse(locEle.Attribute("query_count_before_filter").Value);
}
if (locEle.Attribute("pvalue") != null)
{
fl.PValue = double.Parse(locEle.Attribute("pvalue").Value);
}
foreach (XElement queryEle in locEle.Elements("query"))
{
string qname = queryEle.Attribute("qname").Value;
string loc = queryEle.Attribute("loc").Value;
string key = SAMAlignedLocation.GetKey(qname, loc);
SAMAlignedLocation query = qmmap[key];
FeatureSamLocation fsl = new FeatureSamLocation(fl);
fsl.SamLocation = query;
var attr = queryEle.FindAttribute("overlap");
if (attr == null)
{
fsl.OverlapPercentage = query.OverlapPercentage(fl);
}
else
{
fsl.OverlapPercentage = double.Parse(attr.Value);
}
var nnpm = queryEle.FindAttribute("nnpm");
if (nnpm == null)
{
nnpm = queryEle.FindAttribute("nnmp");
}
if (nnpm != null)
{
fsl.NumberOfNoPenaltyMutation = int.Parse(nnpm.Value);
}
var nmi = queryEle.FindAttribute("nmi");
if (nmi != null)
{
fsl.NumberOfMismatch = int.Parse(nmi.Value);
}
}
}
}
}
qmmap.Clear();
return(result);
}
public static Dictionary<string, Dictionary<string, FeatureItemGroup>> GroupByIdenticalQuery(Dictionary<string, Dictionary<string, FeatureItemGroup>> dic)
{
var featureNames = (from d in dic.Values
from fig in d.Values
from fi in fig
select fi.Name).Distinct().ToList();
var map = new Dictionary<string, FeatureItem>();
foreach (var featureName in featureNames)
{
var item = new FeatureItem() { Name = featureName };
item.Locations.Add(new FeatureLocation());
map[featureName] = item;
}
foreach (var d in dic.Values)
{
foreach (var fig in d.Values)
{
foreach (var fi in fig)
{
var item = map[fi.Name];
item.Locations[0].SamLocations.AddRange(from l in fi.Locations from ll in l.SamLocations select ll);
}
}
}
var groups = map.Values.GroupByIdenticalQuery();
var result = new Dictionary<string, Dictionary<string, FeatureItemGroup>>();
foreach (var sample in dic.Keys)
{
var oldfeatures = (from v in dic[sample].Values
from vv in v
select vv).ToDictionary(m => m.Name);
var newdic = new Dictionary<string, FeatureItemGroup>();
result[sample] = newdic;
foreach (var g in groups)
{
var findfeatures = (from item in g
where oldfeatures.ContainsKey(item.Name)
select oldfeatures[item.Name]).ToList();
if (findfeatures.Count > 0)
{
var newgroup = new FeatureItemGroup();
newgroup.AddRange(findfeatures);
newdic[g.DisplayName] = newgroup;
}
}
}
return result;
}
public static string GetTRNAAminoacid(FeatureItem item)
{
return GetTRNAAminoacid(item.Name);
}
public static string GetTRNACode(FeatureItem item)
{
return GetTRNACode(item.Name);
}