From ce63ceeb3675c0fb81a419d09816b9e3ea9d31c2 Mon Sep 17 00:00:00 2001
From: Zhentian Kai <38963539+ZKai0801@users.noreply.github.com>
Date: Sun, 28 Jun 2020 09:40:24 +0800
Subject: [PATCH 1/2] fix the bug in gene orientation

---
 scripts/make_fusion_genes.py | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/scripts/make_fusion_genes.py b/scripts/make_fusion_genes.py
index 0193daa..a1398b4 100644
--- a/scripts/make_fusion_genes.py
+++ b/scripts/make_fusion_genes.py
@@ -30,7 +30,7 @@ def make_fusion_gene(gene, fw, refflat):
             for line in fh:
                 if gene[0] not in line:
                     continue
-                _, transcript, chrom, _, start, end, _, _, _, exonstart, exonend = line.rstrip("\n").split("\t")
+                _, transcript, chrom, strand, start, end, _, _, _, exonstart, exonend = line.rstrip("\n").split("\t")
                 transcripts[transcript] = (chrom, start, end, exonstart, exonend)
         transcript = get_longest_transcript(transcripts.keys(), refflat)
         chrom, start, end, exonstart, exonend  = transcripts[transcript]
@@ -41,13 +41,15 @@ def make_fusion_gene(gene, fw, refflat):
             for line in fh:
                 if gene[1] not in line:
                     continue
-                _, transcript, chrom, _, start, end, _, _, _, exonstart, exonend = line.rstrip("\n").split("\t")
+                _, transcript, chrom, strand, start, end, _, _, _, exonstart, exonend = line.rstrip("\n").split("\t")
                 break
     
     # write to a file
     header = f">{gene[0]}_{transcript},{chrom}:{start}-{end}\n"
     fw.write(header)
     exons = list(zip(exonstart.split(","), exonend.split(",")))[:-1]
+    if strand == "-":
+        exons = exons[::-1]
     for index, each_exon in enumerate(exons, start=1):
         fw.write(f'{index},{each_exon[0]},{each_exon[1]}\n')
     fw.write("\n")

From ac29e356097c8d1290352316a4ae2f121c8ee76f Mon Sep 17 00:00:00 2001
From: Zhentian Kai <38963539+ZKai0801@users.noreply.github.com>
Date: Sun, 28 Jun 2020 09:46:53 +0800
Subject: [PATCH 2/2] fix the bug of wrong gene/transcript picking

---
 scripts/make_fusion_genes.py | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/scripts/make_fusion_genes.py b/scripts/make_fusion_genes.py
index a1398b4..d80eed1 100644
--- a/scripts/make_fusion_genes.py
+++ b/scripts/make_fusion_genes.py
@@ -28,9 +28,9 @@ def make_fusion_gene(gene, fw, refflat):
         transcripts = {}
         with open(refflat, "r") as fh:
             for line in fh:
-                if gene[0] not in line:
+                cur_gene, transcript, chrom, strand, start, end, _, _, _, exonstart, exonend = line.rstrip("\n").split("\t")
+                if gene[0] != cur_gene:
                     continue
-                _, transcript, chrom, strand, start, end, _, _, _, exonstart, exonend = line.rstrip("\n").split("\t")
                 transcripts[transcript] = (chrom, start, end, exonstart, exonend)
         transcript = get_longest_transcript(transcripts.keys(), refflat)
         chrom, start, end, exonstart, exonend  = transcripts[transcript]
@@ -39,9 +39,9 @@ def make_fusion_gene(gene, fw, refflat):
     elif len(gene) == 2:
         with open(refflat, "r") as fh:
             for line in fh:
-                if gene[1] not in line:
-                    continue
                 _, transcript, chrom, strand, start, end, _, _, _, exonstart, exonend = line.rstrip("\n").split("\t")
+                if gene[1] != transcript:
+                    continue
                 break
     
     # write to a file