protected ReadSummary GetReadSummary(List <FeatureItemGroup> allmapped, HashSet <string> excludeQueries, List <SAMAlignedItem> reads, List <QueryInfo> totalQueries)
{
var result = new ReadSummary();
if (File.Exists(options.CountFile))
{
result.TotalRead = Counts.GetTotalCount();
}
else
{
result.TotalRead = totalQueries.Count;
}
var featureQueries = new HashSet <string>(from fig in allmapped
from fi in fig
from loc in fi.Locations
from sl in loc.SamLocations
select sl.SamLocation.Parent.OriginalQname);
result.FeatureRead = featureQueries.Sum(l => Counts.GetCount(l));
result.ExcludeRead = excludeQueries.Sum(l => Counts.GetCount(l));
result.GenomeRead = (from query in totalQueries
where (!query.Name.Contains(SmallRNAConsts.NTA_TAG) || query.Name.EndsWith(SmallRNAConsts.NTA_TAG))
let originalQname = query.Name.StringBefore(SmallRNAConsts.NTA_TAG)
where !featureQueries.Contains(originalQname) && query.Mismatch == 0 && query.Length >= options.TooShortReadLength
select originalQname).Distinct().Sum(m => Counts.GetCount(m));
if (Counts.ItemMap != null)
{
result.TooShortRead = (from read in Counts.ItemMap.Values
where !featureQueries.Contains(read.Qname) && read.SequenceLength < 20
select read.Count).Sum();
}
else
{
result.TooShortRead = 0;
}
return(result);
}
public override IEnumerable <string> Process()
{
var result = new List <string>();
//read regions
var featureLocations = this.MappedOptions.GetSequenceRegions();
Progress.SetMessage("There are {0} coordinate entries", featureLocations.Count);
if (featureLocations.Count == 0)
{
throw new Exception(string.Format("No coordinate found for {0} in file {1}", options.GtfFeatureName,
options.CoordinateFile));
}
var resultFilename = options.OutputFile;
result.Add(resultFilename);
//parsing reads
List <QueryInfo> totalQueries;
var reads = ParseCandidates(options.InputFile, resultFilename, out totalQueries);
int totalQueryCount;
if (reads.Count == totalQueries.Count && File.Exists(options.CountFile)) //only mapped reads in bam file.
{
totalQueryCount = Counts.GetTotalCount();
}
else
{
totalQueryCount = (from q in totalQueries select q.Name.StringBefore(SmallRNAConsts.NTA_TAG)).Distinct().Sum(m => Counts.GetCount(m));
}
if (reads.Count > 0 && reads[0].Qname.Contains(SmallRNAConsts.NTA_TAG))
{
if (!options.NTA)
{
reads.RemoveAll(m => !m.Qname.EndsWith(SmallRNAConsts.NTA_TAG));
}
}
var totalMappedCount = (from q in reads select q.Qname.StringBefore(SmallRNAConsts.NTA_TAG)).Distinct().Sum(m => Counts.GetCount(m));
Progress.SetMessage("mapping reads to sequence regions...");
MapReadToSequenceRegion(featureLocations, reads);
var featureReadCount = reads.Where(m => m.Locations.Any(n => n.Features.Count > 0)).Sum(m => m.QueryCount);
Console.WriteLine("feature reads = {0}", featureReadCount);
var mappedItems = featureLocations.GroupByName();
mappedItems.RemoveAll(m => m.GetEstimatedCount() == 0);
mappedItems.ForEach(m => m.CombineLocations());
var mappedGroups = mappedItems.GroupByIdenticalQuery();
//group by miRNA name
if (!options.NoMappedFile)
{
Progress.SetMessage("output mapping details...");
var mappedfile = resultFilename + ".mapped.xml";
new FeatureItemGroupXmlFormat().WriteToFile(mappedfile, mappedGroups);
result.Add(mappedfile);
}
Progress.SetMessage("write result ...");
mappedGroups.Sort((m1, m2) => m2.GetEstimatedCount().CompareTo(m1.GetEstimatedCount()));
new FeatureItemGroupCountWriter().WriteToFile(resultFilename, mappedGroups);
if (options.ExportLengthDistribution)
{
var disfile = resultFilename + ".length";
new FeatureItemGroupReadLengthWriter().WriteToFile(disfile, mappedGroups);
result.Add(disfile);
}
if (options.ExportSequenceCount)
{
var seqfile = resultFilename + ".seqcount";
new FeatureItemGroupSequenceWriter().WriteToFile(seqfile, mappedGroups);
result.Add(seqfile);
}
if (options.UnmappedFastq)
{
Progress.SetMessage("output unmapped query...");
var unmappedFile = Path.ChangeExtension(resultFilename, ".unmapped.fastq.gz");
var except = new HashSet <string>(from r in reads
where r.Locations.Count > 0
select r.Qname);
if (File.Exists(options.FastqFile))
{
new FastqExtractorFromFastq {
Progress = Progress
}.Extract(options.FastqFile, unmappedFile, except, options.CountFile);
}
else
{
new FastqExtractorFromBam()
{
Progress = Progress
}.Extract(options.InputFile, unmappedFile, except, options.CountFile);
}
result.Add(unmappedFile);
}
Progress.SetMessage("summarizing ...");
var infoFile = Path.ChangeExtension(resultFilename, ".info");
using (var sw = new StreamWriter(infoFile))
{
sw.WriteLine("#file\t{0}", options.InputFile);
sw.WriteLine("#coordinate\t{0}", options.CoordinateFile);
sw.WriteLine("#minLength\t{0}", options.MinimumReadLength);
sw.WriteLine("#maxMismatchCount\t{0}", options.MaximumMismatch);
if (File.Exists(options.CountFile))
{
sw.WriteLine("#countFile\t{0}", options.CountFile);
}
sw.WriteLine("TotalReads\t{0}", totalQueryCount);
sw.WriteLine("MappedReads\t{0}", totalMappedCount);
sw.WriteLine("MultipleMappedReads\t{0}", reads.Where(m => m.Locations.Count > 1).Sum(m => m.QueryCount));
sw.WriteLine("FeatureReads\t{0}", featureReadCount);
}
result.Add(infoFile);
Progress.End();
return(result);
}