better gtf file support

figeno/cli/cli.py

@@ -2,7 +2,7 @@
 
 from figeno.cli import gui, init,make
 
-__version__ = "1.4.3"
+__version__ = "1.4.4"
 
 def main():
     parser = ArgumentParser("figeno",formatter_class=ArgumentDefaultsHelpFormatter)

figeno/genes.py

@@ -179,7 +179,7 @@ def read_transcripts_names(file,gene_names):
 
 
 
-def read_transcripts(file,chr=None,start=None,end=None,gene_names="auto",collapsed=True,only_protein_coding=True):
+def read_transcripts(file,chr=None,start=None,end=None,gene_names="auto",collapsed=True,only_protein_coding=True,warnings=[]):
     """Wrapper depending on the file type"""
     if gene_names =="auto" or gene_names=="": gene_names = None
     if gene_names is not None: 
@@ -190,7 +190,7 @@ def read_transcripts(file,chr=None,start=None,end=None,gene_names="auto",collaps
     elif file.endswith("gff3") or file.endswith("gff3.gz"):
         return read_genes_gff3(file,chr,start,end,gene_names,collapsed=collapsed,only_protein_coding=only_protein_coding) 
     elif file.endswith("gtf") or file.endswith("gtf.gz"):
-        return read_genes_gtf(file,chr,start,end,gene_names,collapsed=collapsed,only_protein_coding=only_protein_coding) 
+        return read_genes_gtf(file,chr,start,end,gene_names,collapsed=collapsed,only_protein_coding=only_protein_coding,warnings=warnings) 
     else:
         raise KnownException("The extension for genes file "+str(file)+" was not recognized. The filename must end with .txt(.gz) for RefSeq format, .gtf(.gz) for gtf or .gff3(.gz) for gff3.")
 
@@ -267,7 +267,7 @@ def read_genes_refseq(file,chr=None,start=None,end=None,gene_names=None,collapse
 
 
 
-def read_genes_gtf(gtf_file,chr=None,start=None,end=None,gene_names=None,collapsed=True,only_protein_coding=True):
+def read_genes_gtf(gtf_file,chr=None,start=None,end=None,gene_names=None,collapsed=True,only_protein_coding=True,warnings=[]):
     if gene_names is not None: gene_names=gene_names.upper()
     transcripts={}
     name2exons={}
@@ -294,8 +294,16 @@ def read_genes_gtf(gtf_file,chr=None,start=None,end=None,gene_names=None,collaps
                 elif x.lstrip(" ").startswith("transcript_name"):
                     x = x[x.find("\"")+1:]
                     transcript_name = x[:x.find("\"")]
+                elif transcript_name=="" and x.lstrip(" ").startswith("transcript_id"):
+                    x = x[x.find("\"")+1:]
+                    transcript_name = x[:x.find("\"")]
             if (gene_names is not None) and (not gene_name.upper() in gene_names): continue
-            if only_protein_coding and (not "protein_coding" in linesplit[8]): continue
+            if only_protein_coding and (not "protein_coding" in linesplit[8]):
+                if ("gene_biotype" in linesplit[8]): continue
+                else: 
+                    warn_message="Could not filter for protein coding genes because this information was not provided in the gtf file."
+                    if not warn_message in warnings: warnings.append(warn_message) 
+
             if collapsed: name = gene_name
             else: name=transcript_name
             transcript = Transcript(gene_name,linesplit[0].lstrip("chr"),transcript_start,transcript_end,transcript_orientation,[])
@@ -310,7 +318,10 @@ def read_genes_gtf(gtf_file,chr=None,start=None,end=None,gene_names=None,collaps
                 if x.lstrip(" ").startswith("transcript_name"):
                     x = x[x.find("\"")+1:]
                     transcript_name = x[:x.find("\"")]
-                if x.lstrip(" ").startswith("gene_name"):
+                elif transcript_name=="" and x.lstrip(" ").startswith("transcript_id"):
+                    x = x[x.find("\"")+1:]
+                    transcript_name = x[:x.find("\"")]
+                elif x.lstrip(" ").startswith("gene_name"):
                     x = x[x.find("\"")+1:]
                     gene_name = x[:x.find("\"")]
             if (gene_names is not None) and (not gene_name in gene_names): continue

figeno/gui/package.json

@@ -1,6 +1,6 @@
 {
   "name": "figeno",
-  "version": "1.4.3",
+  "version": "1.4.4",
   "private": true,
   "homepage": "./",
   "dependencies": {

figeno/track_genes.py

@@ -29,7 +29,7 @@ def __init__(self,reference="custom",genes_file="",style="default",collapsed=Tru
     def draw(self, regions, box ,hmargin=0,warnings=[]):
         boxes = split_box(box,regions,hmargin)
         self.margin_between_genes = 1.5*np.sum([abs(r[0].end-r[0].start) for r in regions]) / abs(box["right"]-box["left"]) # in bp
-        lines_regions = self.read_transcripts_lines_regions(regions)
+        lines_regions = self.read_transcripts_lines_regions(regions,warnings=warnings)
         max_region_size=0
         for i in range(len(regions)):
             max_region_size=max(max_region_size,abs(regions[i][0].end-regions[i][0].start))
@@ -175,7 +175,7 @@ def draw_title(self,box):
                 box["ax"].text(box["left"] - 1.0,(box["top"]+box["bottom"])/2,
                             self.label,rotation=rotation,horizontalalignment="right",verticalalignment="center",fontsize=7*self.fontscale)
 
-    def read_transcripts_lines_regions(self,regions):
+    def read_transcripts_lines_regions(self,regions,warnings=[]):
         regions = [reg[0] for reg in regions]
         lines_regions=[]
         max_nlines=0
@@ -188,7 +188,7 @@ def read_transcripts_lines_regions(self,regions):
                     raise KnownException("When using a custom reference genome, you have to provide a genes file if you want to display a genes track. See https://figeno.readthedocs.io/en/latest/content/describe_figure.html#general for the format of this file.")
             else:
                 transcripts = read_transcripts(self.genes_file,region.chr,region.start,region.end,self.genes,
-                                               collapsed=self.collapsed,only_protein_coding=self.only_protein_coding)
+                                               collapsed=self.collapsed,only_protein_coding=self.only_protein_coding,warnings=warnings)
 
             # Assign transcripts to lines, so that they do not overlap
             transcripts= sorted(transcripts,key=lambda x:x.start)

pyproject.toml

@@ -9,7 +9,7 @@ packages = ["figeno", "figeno.data", "figeno.cli", "figeno.gui"]
 
 [project]
 name = 'figeno'
-version = "1.4.3"
+version = "1.4.4"
 description = 'Package for generating genomics figures.'
 readme = 'README.md'
 authors = [