public void Execute()
{
var validTss = this.TranscriptExpression
.Select(x => new
{
Values = x.Value.Values,
TissueExpressionData = new Analyses.MapBuilderData.TissueExpressionData
{
Tss = x.Key,
Location = this.TranscriptLocations[x.Key],
Expression = x.Value,
}
})
.Where(x => this.IsValidExpressionData(x.TissueExpressionData) && CorrelationMapBuilder.IsTwoFoldChange(x.Values))
.ToList();
string stem = string.Join("_", new string[]
{
"TssSet",
this.Tissue,
this.RnaSource,
this.HistoneName,
}
.Concat(this.OmittedTissues != null ? this.OmittedTissues : new string[] {}));
string file = string.Format("../temp/results/TssSets/{0}.nsv", stem);
Tables.ToNamedNsvFile(file, validTss.Select(x => x.TissueExpressionData.Tss));
if (this.AnnotationFileName != null)
{
var annotation = new GtfExpressionFile(
GtfExpressionFile.ExpressionTypeFromString(this.RnaSource),
this.AnnotationFileName);
var tssToGene = annotation.Transcripts.ToDictionary(x => x.Key, x => x.Value.AlternateName);
string geneStem = string.Join("_", new string[]
{
"GeneSet",
this.Tissue,
this.RnaSource,
this.HistoneName,
}
.Concat(this.OmittedTissues != null ? this.OmittedTissues : new string[] { }));
string geneFile = string.Format("../temp/results/TssSets/{0}.nsv", geneStem);
Tables.ToNamedNsvFile(geneFile, validTss
.Where(x => tssToGene.ContainsKey(x.TissueExpressionData.Tss))
.ToLookup(x => tssToGene[x.TissueExpressionData.Tss], x => x)
.Select(x => x.Key));
}
}
/// <summary>
/// Convert the map to genes.
/// </summary>
public void Convert()
{
MapLinkFilter filter = new MapLinkFilter
{
MaximumLinkLength = this.MaxRange,
ConfidenceThreshold = this.PvalueThreshold,
LinkTypeFilter = MapLinkFilter.LinkType.Any,
};
TssRegulatoryMap map = TssRegulatoryMap.LoadMap(this.MapFileName, filter);
var expression = new GtfExpressionFile(
GtfExpressionFile.ExpressionTypeFromString(this.RnaSource),
this.AnnotationFileName);
var geneMap = map.ConvertToGenes(IUnknown.QueryInterface <IExpressionData>(expression));
foreach (var link in geneMap.Links)
{
var geneLocation = expression.Transcripts[link.TranscriptName];
string[] lineData = new string[]
{
geneLocation.Chromosome,
geneLocation.Start.ToString(),
geneLocation.End.ToString(),
link.GeneName,
"NA",
geneLocation.Strand,
link.LocusName,
link.Correlation.ToString(),
link.ConfidenceScore.ToString(),
link.LinkLength.ToString(),
this.HistoneName,
link.TranscriptName,
};
Console.WriteLine(string.Join("\t", lineData));
}
}
public void Convert()
{
var contacts = Helpers.GetFileDataLines(this.ContactFileName, true).Select(line =>
{
var fields = line.Split('\t');
var pair = fields[6];
var ids = pair.Split('_');
return(new
{
Pair = pair,
Id1 = ids[0],
Chr1 = fields[0],
Start1 = int.Parse(fields[1]),
End1 = int.Parse(fields[2]),
Id2 = ids[1],
Chr2 = fields[3],
Start2 = int.Parse(fields[4]),
End2 = int.Parse(fields[5]),
Score = double.Parse(fields[7]),
});
}).ToList();
var locations = contacts.Select(x => new Location[]
{
new Location
{
Name = x.Id1, Chromosome = x.Chr1, Start = x.Start1, End = x.End1, AlternateName = x.Id2, Score = x.Score
},
new Location
{
Name = x.Id2, Chromosome = x.Chr2, Start = x.Start2, End = x.End2, AlternateName = x.Id1, Score = x.Score
},
}).SelectMany(x => x)
.ToList();
var uniqueLocations = locations
.ToLookup(x => x.Name)
.ToDictionary(x => x.Key, x => x.First());
var contactMap = locations
.ToLookup(x => x.Name, x => x)
.ToDictionary(x => x.Key, x => x
.ToDictionary(y => y.AlternateName, y => uniqueLocations[y.AlternateName]));
var indexedLocations = new BedFile(uniqueLocations);
var LocusFile = new BedFile(this.LocusFileName, BedFile.Bed3Layout);
var annotationSet = new GtfExpressionFile(GtfExpressionFile.ExpressionType.Cage, this.AnnotationFileName);
var LocusOverlaps = indexedLocations.Locations.ToDictionary(
x => x.Key,
x => TRFScorer.GetOverlaps(
x.Value,
LocusFile.ChromosomeIndexedLocations,
BedFile.IndexSize,
LocusFile.MaxLocationSize[x.Value.Chromosome]));
var tssOverlaps = indexedLocations.Locations.ToDictionary(
x => x.Key,
x => TRFScorer.GetOverlaps(
x.Value,
annotationSet.ChromosomeIndexedLocations,
BedFile.IndexSize,
annotationSet.MaxLocationSize[x.Value.Chromosome],
Location.OverlapsDirectionalStart));
var links = contactMap
.Where(x => tssOverlaps[x.Key] != null)
.ToDictionary(x => x.Key, x => x.Value.Where(y => LocusOverlaps[y.Key] != null))
.Where(x => x.Value.Any())
.Select(x => tssOverlaps[x.Key].Select(y => new {
Id1 = x.Key,
Tss = y.Name,
LocusList = x.Value
.Select(z => LocusOverlaps[z.Key]
.Select(a => new { Name = a.Name, Id2 = z.Key }))
.SelectMany(z => z).ToList()
})
.Select(y => y.LocusList.Select(z =>
{
var LocusLocation = LocusFile.Locations[z.Name];
var tssStart = annotationSet.Transcripts[y.Tss].DirectionalStart;
var linkLength = LocusLocation.End < tssStart ? LocusLocation.End - tssStart : LocusLocation.Start - tssStart;
return(new MapLink
{
TranscriptName = y.Tss,
TssName = y.Tss,
LocusName = z.Name,
ConfidenceScore = contactMap[y.Id1][z.Id2].Score,
LinkLength = linkLength,
});
}))
.SelectMany(y => y))
.SelectMany(x => x)
.ToList();
var convertedMap = new TssRegulatoryMap(links);
NullMapBuilder.WriteMap(convertedMap, annotationSet.Locations, "Contact", this.MapFileName);
}
