/// <summary>
/// Reads gene model features into data structures contained within this library
/// </summary>
/// <param name="geneModelFile"></param>
public void ReadGeneFeatures(string geneModelFile)
{
foreach (ISequence chromFeatures in SimplerParse(geneModelFile))
{
Chromosome chrom = Genome.Chromosomes.FirstOrDefault(x => x.FriendlyName == chromFeatures.ID);
if (chrom == null)
{
continue;
}
chromFeatures.Metadata.TryGetValue("features", out object f);
List <MetadataListItem <List <string> > > features = f as List <MetadataListItem <List <string> > >;
for (int i = 0; i < features.Count; i++)
{
MetadataListItem <List <string> > feature = features[i];
long.TryParse(feature.SubItems["start"][0], out long start);
long.TryParse(feature.SubItems["end"][0], out long end);
var attributes = SplitAttributes(feature.FreeText);
if (feature.FreeText.Contains('='))
{
ProcessGff3Feature(feature, start, end, chrom, attributes);
}
else
{
ProcessGtfFeature(feature, start, end, chrom, attributes);
}
}
}
if (currentTranscript != null)
{
Transcript.SetRegions(currentTranscript);
currentTranscript.FrameCorrection();
}
CreateIntergenicRegions();
// possibly check transcript sanity here with Parallel.ForEach(Genes.SelectMany(g => g.Transcripts).ToList(), t => t.SanityCheck());
GenomeForest.Build();
}
/// <summary>
/// Processes a feature from a GTF gene model file.
/// </summary>
/// <param name="feature"></param>
/// <param name="oneBasedStart"></param>
/// <param name="oneBasedEnd"></param>
/// <param name="chrom"></param>
/// <param name="attributes"></param>
public void ProcessGtfFeature(MetadataListItem <List <string> > feature, long oneBasedStart, long oneBasedEnd, Chromosome chrom, Dictionary <string, string> attributes)
{
bool hasGeneId = attributes.TryGetValue("gene_id", out string geneId);
bool hasTranscriptId = attributes.TryGetValue("transcript_id", out string transcriptId);
bool hasProteinId = attributes.TryGetValue("protein_id", out string proteinId);
bool hasExonId = attributes.TryGetValue("exon_id", out string exonId);
bool hasSource = feature.SubItems.TryGetValue("source", out List <string> sourceish);
bool hasStrand = feature.SubItems.TryGetValue("strand", out List <string> strandish);
bool hasFrame = feature.SubItems.TryGetValue("frame", out List <string> framey);
string source = hasSource ? sourceish[0] : "";
if (!hasStrand)
{
return;
} // strand is a required to do anything in this program
string strand = strandish[0];
int frame = 0;
if (hasFrame)
{
int.TryParse(framey[0], out frame);
}
// Trim prefixes from the IDs
string genePrefix = "gene:";
string transcriptPrefix = "transcript:";
if (hasGeneId && geneId.StartsWith(genePrefix))
{
string newGeneId = geneId.Substring(genePrefix.Length);
feature.FreeText.Replace(geneId, newGeneId);
geneId = newGeneId;
}
if (hasTranscriptId && transcriptId.StartsWith(transcriptPrefix))
{
string newTranscriptId = transcriptId.Substring(transcriptPrefix.Length);
feature.FreeText.Replace(transcriptId, newTranscriptId);
transcriptId = newTranscriptId;
}
if (hasProteinId && proteinId.StartsWith(transcriptPrefix))
{
proteinId = proteinId.Substring(transcriptPrefix.Length); // transcript id is used for protein id sometimes
}
// Catch the transcript features before they go by if available, i.e. if the file doesn't just have exons
if (feature.Key == "transcript" && (currentTranscript == null || hasTranscriptId && transcriptId != currentTranscript.ID))
{
if (currentGene == null || hasGeneId && geneId != currentGene.ID)
{
currentGene = new Gene(geneId, chrom, source, strand, oneBasedStart, oneBasedEnd, feature);
Genes.Add(currentGene);
GenomeForest.Add(currentGene);
}
currentTranscript = new Transcript(transcriptId, currentGene, source, strand, oneBasedStart, oneBasedEnd, null, null, feature);
currentGene.Transcripts.Add(currentTranscript);
GenomeForest.Add(currentTranscript);
}
if (feature.Key == "exon" || feature.Key == "CDS")
{
if (currentGene == null || hasGeneId && geneId != currentGene.ID)
{
currentGene = new Gene(geneId, chrom, source, strand, oneBasedStart, oneBasedEnd, feature);
Genes.Add(currentGene);
GenomeForest.Add(currentGene);
}
if (currentTranscript == null || hasTranscriptId && transcriptId != currentTranscript.ID)
{
if (currentTranscript != null)
{
Transcript.SetRegions(currentTranscript);
currentTranscript.FrameCorrection();
}
currentTranscript = new Transcript(transcriptId, currentGene, source, strand, oneBasedStart, oneBasedEnd, null, null, feature);
currentGene.Transcripts.Add(currentTranscript);
GenomeForest.Add(currentTranscript);
}
if (feature.Key == "exon")
{
ISequence exon_dna = chrom.Sequence.GetSubSequence(oneBasedStart - 1, oneBasedEnd - oneBasedStart + 1);
Exon exon = new Exon(currentTranscript, currentTranscript.IsStrandPlus() ? exon_dna : exon_dna.GetReverseComplementedSequence(),
source, oneBasedStart, oneBasedEnd, chrom.Sequence.ID, strand, null, feature);
if (exon.Length() > 0)
{
currentTranscript.Exons.Add(exon);
}
}
else if (feature.Key == "CDS")
{
CDS cds = new CDS(currentTranscript, chrom.Sequence.ID, source, strand, oneBasedStart, oneBasedEnd, null, frame);
if (hasProteinId)
{
currentTranscript.ProteinID = proteinId;
}
if (cds.Length() > 0)
{
currentTranscript.CodingDomainSequences.Add(cds);
}
}
else
{ // nothing to do
}
}
}
/// <summary>
/// Processes a feature from a GFF3 gene model file.
/// </summary>
/// <param name="feature"></param>
/// <param name="oneBasedStart"></param>
/// <param name="oneBasedEnd"></param>
/// <param name="chrom"></param>
/// <param name="attributes"></param>
public void ProcessGff3Feature(MetadataListItem <List <string> > feature, long oneBasedStart, long oneBasedEnd, Chromosome chrom, Dictionary <string, string> attributes)
{
bool hasGeneId = attributes.TryGetValue("gene_id", out string geneId);
bool hasTranscriptId = attributes.TryGetValue("transcript_id", out string transcriptId);
bool hasExonId = attributes.TryGetValue("exon_id", out string exonId);
bool hasProteinId = attributes.TryGetValue("protein_id", out string proteinId);
bool hasSource = feature.SubItems.TryGetValue("source", out List <string> sourceish); // false if empty ("." in GFF format)
bool hasStrand = feature.SubItems.TryGetValue("strand", out List <string> strandish); // false if empty ("." in GFF format)
bool hasFrame = feature.SubItems.TryGetValue("frame", out List <string> framey); // false if empty ("." in GFF format)
string source = hasSource ? sourceish[0] : "";
if (!hasStrand)
{
return;
} // strand is a required to do anything in this program
string strand = strandish[0];
int frame = 0;
if (hasFrame)
{
int.TryParse(framey[0], out frame);
}
if (hasGeneId && (currentGene == null || hasGeneId && geneId != currentGene.ID))
{
currentGene = new Gene(geneId, chrom, source, strand, oneBasedStart, oneBasedEnd, feature);
Genes.Add(currentGene);
GenomeForest.Add(currentGene);
}
if (hasTranscriptId && (currentTranscript == null || hasTranscriptId && transcriptId != currentTranscript.ID))
{
if (currentTranscript != null)
{
Transcript.SetRegions(currentTranscript);
currentTranscript.FrameCorrection();
}
currentTranscript = new Transcript(transcriptId, currentGene, source, strand, oneBasedStart, oneBasedEnd, null, null, feature);
currentGene.Transcripts.Add(currentTranscript);
GenomeForest.Add(currentTranscript);
}
if (hasExonId)
{
ISequence exon_dna = chrom.Sequence.GetSubSequence(oneBasedStart - 1, oneBasedEnd - oneBasedStart + 1);
Exon exon = new Exon(currentTranscript, currentTranscript.IsStrandPlus() ? exon_dna : exon_dna.GetReverseComplementedSequence(),
source, oneBasedStart, oneBasedEnd, chrom == null ? "" : chrom.ChromosomeID, strand, null, feature);
if (exon.Length() > 0)
{
currentTranscript.Exons.Add(exon);
}
}
else if (hasProteinId)
{
CDS cds = new CDS(currentTranscript, chrom.Sequence.ID, source, strand, oneBasedStart, oneBasedEnd, null, frame);
if (cds.Length() > 0)
{
currentTranscript.CodingDomainSequences.Add(cds);
currentTranscript.ProteinID = proteinId;
}
}
else // nothing to do
{
}
}
