Account for stringtie multigene transcript artefacts

.gitlab-ci.yml

@@ -12,6 +12,8 @@ variables:
               --direct_rna --minimap2_index_opts '-k 15' --sample_sheet ${CI_PROJECT_NAME}/data/differential_expression/sample_sheet.csv \
               -c ${CI_PROJECT_NAME}/data/demo.nextflow.config"
               CI_FLAVOUR: "new"
+              PYTEST_CONTAINER_NAME: "wf-common"
+              PYTEST_CONTAINER_CONFIG_KEY: "common_sha"
 
 macos-run:
   # Let's avoid those ARM64 runners for now
@@ -146,7 +148,7 @@ docker-run:
                 --ref_genome ${CI_PROJECT_NAME}/data/differential_expression_ncbi/GCF_000001405.40_GRCh38.p14_genomic.fna.gz \
                 --ref_annotation ${CI_PROJECT_NAME}/data/differential_expression_ncbi/GCF_000001405.40_GRCh38.p14_genomic.gff.gz \
                 --direct_rna --ref_transcriptome ${CI_PROJECT_NAME}/data/differential_expression_ncbi/GCF_000001405.40_GRCh38.p14_rna.fna.gz \
-                --transcriptome_assembly false --sample_sheet test_data/sample_sheet.csv \
+                --sample_sheet test_data/sample_sheet.csv \
                 -c ${CI_PROJECT_NAME}/data/demo.nextflow.config"
               NF_IGNORE_PROCESSES: >
                 preprocess_reads,faidx,gz_faidx,merge_transcriptomes,assemble_transcripts,
@@ -160,7 +162,7 @@ docker-run:
                 --ref_genome ${CI_PROJECT_NAME}/data/differential_expression/Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz \
                 --ref_annotation  ${CI_PROJECT_NAME}/data/differential_expression/Homo_sapiens.GRCh38.109.gtf.gz \
                 --direct_rna --ref_transcriptome  ${CI_PROJECT_NAME}/data/differential_expression/Homo_sapiens.GRCh38.cdna.all.fa.gz \
-                --transcriptome_assembly false --sample_sheet test_data/sample_sheet.csv \
+                --sample_sheet test_data/sample_sheet.csv \
                 -c ${CI_PROJECT_NAME}/data/demo.nextflow.config"
               NF_IGNORE_PROCESSES: >
                 preprocess_reads,faidx,gz_faidx,merge_transcriptomes,assemble_transcripts,
@@ -174,7 +176,7 @@ docker-run:
                 --ref_genome  ${CI_PROJECT_NAME}/data/differential_expression_mouse/GRCm39.genome.fa.gz \
                 --ref_annotation  ${CI_PROJECT_NAME}/data/differential_expression_mouse/gencode.vM33.annotation.gtf \
                 --direct_rna --ref_transcriptome  ${CI_PROJECT_NAME}/data/differential_expression_mouse/gencode.vM33.transcripts.fa.gz \
-                --transcriptome_assembly false --sample_sheet  ${CI_PROJECT_NAME}/data/differential_expression_mouse/sample_sheet.csv \
+                --sample_sheet  ${CI_PROJECT_NAME}/data/differential_expression_mouse/sample_sheet.csv \
                 -c  ${CI_PROJECT_NAME}/data/demo.nextflow.config"
               NF_IGNORE_PROCESSES: >
                 preprocess_reads,faidx,gz_faidx,merge_transcriptomes,assemble_transcripts,decompress_annotation,
@@ -215,7 +217,7 @@ docker-run:
                 --ref_genome ${CI_PROJECT_NAME}/data/differential_expression_ncbi/GCF_000001405.40_GRCh38.p14_genomic.fna.gz \
                 --ref_annotation ${CI_PROJECT_NAME}/data/differential_expression_ncbi/GCF_000001405.40_GRCh38.p14_genomic.gff.gz \
                 --direct_rna --ref_transcriptome ${CI_PROJECT_NAME}/data/differential_expression_ncbi/GCF_000001405.40_GRCh38.p14_rna.fna.gz \
-                --transcriptome_assembly false --sample_sheet test_data/sample_sheet.csv \
+                --sample_sheet test_data/sample_sheet.csv \
                 --igv \
                 -c ${CI_PROJECT_NAME}/data/demo.nextflow.config"
               NF_IGNORE_PROCESSES: >

CHANGELOG.md

@@ -4,6 +4,10 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [Unreleased]
+### Fixed
+- Bug that led to incorrect gene_id being assigned in the DE plots.
+
 ## [v1.5.0]
 ### Updated
 - Workflow report updated to use `ezcharts`.

bin/workflow_glue/de_plots.py

@@ -1,6 +1,5 @@
 #!/usr/bin/env python
 """Create de report section."""
-
 import os
 
 from dominate.tags import h5, p
@@ -52,7 +51,7 @@ def create_summary_table(df):
         avg_acc, avg_mapq])
 
 
-def dexseq_section(dexseq_file, id_dic, pval_thresh):
+def dexseq_section(dexseq_file, tr_id_to_gene_name, tr_id_to_gene_id, pval_thresh):
     """Add gene isoforms table and plot."""
     h5("Differential Isoform usage")
     p("""Table showing gene isoforms, ranked by adjusted
@@ -64,9 +63,17 @@ def dexseq_section(dexseq_file, id_dic, pval_thresh):
 
     dexseq_results = pd.read_csv(dexseq_file, sep='\t')
     dexseq_results.index.name = "gene_id:transcript_id"
-    # Replace gene id with more useful gene name where possible
+
+    # Replace any occurrences of stringtie-generated MSTRG gene ids with
+    # reference gene_ids.
     dexseq_results.index = dexseq_results.index.map(
-        lambda x: str(id_dic.get(x.split(':')[0])) + ':' + str(x.split(':')[1]))
+        lambda ge_tr: str(  # lookup gene_id from transcript_id [1]
+            f"{tr_id_to_gene_id.get(ge_tr.split(':')[1])}: {str(ge_tr.split(':')[1])}")
+    )
+
+    # Add gene name column.
+    dexseq_results.insert(0, "gene_name", dexseq_results.index.map(
+        lambda x: tr_id_to_gene_name.get(x.split(':')[1])))
 
     DataTable.from_pandas(
         dexseq_results.sort_values(by='pvalue', ascending=True), use_index=True)
@@ -100,13 +107,12 @@ def dexseq_section(dexseq_file, id_dic, pval_thresh):
     EZChart(plot)
 
 
-def dtu_section(dtu_file, gt_dic, ge_dic):
+def dtu_section(dtu_file, txid_to_gene_name):
     """Plot dtu section."""
     dtu_results = pd.read_csv(dtu_file, sep='\t')
     dtu_results["gene_name"] = dtu_results["txID"].apply(
-        lambda x: gt_dic.get(x))
+        lambda x: txid_to_gene_name.get(x))
-    dtu_results["geneID"] = dtu_results["geneID"].apply(
+
-        lambda x: ge_dic.get(x))
     dtu_pvals = dtu_results.sort_values(by='gene', ascending=True)
     raw("""Table showing gene and transcript identifiers
     and their FDR-corrected (False discovery rate - Benjamini-Hochberg) probabilities
@@ -120,11 +126,10 @@ def dtu_section(dtu_file, gt_dic, ge_dic):
     raw("""View dtu_plots.pdf file to see plots of differential isoform usage""")
 
 
-def dge_section(dge_file, ids_dic, pval_thresh):
+def dge_section(df, pval_thresh):
     """Create DGE table and MA plot."""
     h5("Differential gene expression")
-    dge_results = pd.read_csv(dge_file, sep='\t')
+    df[['logFC', 'logCPM', 'F']] = df[
-    dge_results[['logFC', 'logCPM', 'F']] = dge_results[
         ['logFC', 'logCPM', 'F']].round(2)
 
     p("""Table showing the genes from the edgeR analysis.
@@ -134,10 +139,9 @@ def dge_section(dge_file, ids_dic, pval_thresh):
     This table has not been
     filtered for genes that satisfy statistical or magnitudinal thresholds""")
 
-    dge_results.index = dge_results.index.map(lambda x: ids_dic.get(x))
+    df = df.sort_values('FDR', ascending=True)
-    dge_results = dge_results.sort_values('FDR', ascending=True)
+    df.index.name = 'gene_id'
-    dge_results.index.name = 'Transcript'
+    DataTable.from_pandas(df, use_index=True)
-    DataTable.from_pandas(dge_results, use_index=True)
 
     h5("Results of the edgeR Analysis.")
 
@@ -150,15 +154,13 @@ def dge_section(dge_file, ids_dic, pval_thresh):
     (False discovery rate - Benjamini-Hochberg) p-value thresholds
     defined are shaded as 'Up-' or 'Down-' regulated.
     """)
-
+    df['sig'] = None
-    dge = pd.read_csv(dge_file, sep="\t")
+    df.loc[(df["logFC"] > 0) & (df['PValue'] < pval_thresh), 'sig'] = 'up'
-    dge['sig'] = None
+    df.loc[(df["logFC"] <= 0) & (df['PValue'] < pval_thresh), 'sig'] = 'down'
-    dge.loc[(dge["logFC"] > 0) & (dge['PValue'] < pval_thresh), 'sig'] = 'up'
+    df.loc[(df["PValue"] >= pval_thresh), 'sig'] = 'not_sig'
-    dge.loc[(dge["logFC"] <= 0) & (dge['PValue'] < pval_thresh), 'sig'] = 'down'
-    dge.loc[(dge["PValue"] >= pval_thresh), 'sig'] = 'not_sig'
 
     plot = scatterplot(
-        data=dge, x='logCPM', y='logFC', hue='sig',
+        data=df, x='logCPM', y='logFC', hue='sig',
         palette=['#E32636', '#7E8896', '#0A22DE'],
         hue_order=['up', 'not_sig', 'down'], marker='circle')
     plot._fig.x_range.start = 10
@@ -188,53 +190,58 @@ def salmon_table(salmon_counts):
 
 
 def get_translations(gtf):
-    """Create dict with gene_name and gene_references."""
+    """Create gene_and transcript id mappings.
+
+    Annotation can be stringtie-generated (GTF) or from the input
+    reference annotation (GTF or GFF3) and the various attributes can differ
+    """
     with open(gtf) as fh:
-        gene_txid = {}
+        txid_to_gene_name = {}
-        gene_geid = {}
+        gid_to_gene_name = {}
-        geid_gname = {}
+        tx_id_to_gene_id = {}
 
         def get_feature(row, feature):
             return row.split(feature)[1].split(
                 ";")[0].replace('=', '').replace("\"", "").strip()
 
-        for i in fh:
+        for gff_entry in fh:
-            if i.startswith("#"):
+            # Process transcripts features only
+            if gff_entry.startswith("#") or gff_entry.split('\t')[2] != 'transcript':
                 continue
             # Different gtf/gff formats contain different attributes
             # and different formating (eg. gene_name="xyz" or gene_name "xyz")
-            gene_name = None
+            gene_name = gene_id = transcript_id = 'unknown'
-            for var_name in ["gene_name", "gene_id", "gene"]:
-                if var_name in i:
-                    gene_name = get_feature(i, var_name)
-                    break
 
-            if 'ref_gene_id' in i:
+            if 'ref_gene_id' in gff_entry:
-                gene_reference = get_feature(i, 'ref_gene_id')
+                # Favour ref_gene_id over gene_id. The latter can be multi-locus merged
-            elif 'gene_id' in i:
+                # genes from stringtie
-                gene_reference = get_feature(i, 'gene_id')
+                gene_id = get_feature(gff_entry, 'ref_gene_id')
+            elif 'gene_id' in gff_entry:
+                gene_id = get_feature(gff_entry, 'gene_id')
             else:
-                gene_reference = gene_name
+                gene_id = get_feature(gff_entry, 'gene')
-            if 'transcript_id' in i:
+
-                transcript_id = get_feature(i, 'transcript_id')
+            if 'transcript_id' in gff_entry:
+                transcript_id = get_feature(gff_entry, 'transcript_id')
+
+            if 'gene_name' in gff_entry:
+                gene_name = get_feature(gff_entry, 'gene_name')
             else:
-                transcript_id = "unknown"
+                # Fallback to gene_id if gene_name is not present
-            if 'gene_id' in i:
+                gene_name = gene_id
-                gene_id = get_feature(i, 'gene_id')
+
-            else:
+            txid_to_gene_name[transcript_id] = gene_name
-                gene_id = gene_name
+            tx_id_to_gene_id[transcript_id] = gene_id
-            gene_txid[transcript_id] = gene_name
+            gid_to_gene_name[gene_id] = gene_name
-            gene_geid[gene_id] = gene_reference
+    return txid_to_gene_name, tx_id_to_gene_id, gid_to_gene_name
-            geid_gname[gene_reference] = gene_name
-    return gene_txid, gene_geid, geid_gname
 
 
 def de_section(
-        stringtie, dge, dexseq, dtu,
+        annotation, dge, dexseq, dtu,
         tpm, report, filtered, unfiltered,
         gene_counts, aln_stats_dir, pval_threshold=0.01):
     """Differential expression sections."""
-    with report.add_section("Differential expression", "DE"):
+    with (report.add_section("Differential expression", "DE")):
 
         p("""This section shows differential gene expression
         and differential isoform usage. Salmon was used to
@@ -256,39 +263,45 @@ def de_section(
         DataTable.from_pandas(alignment_summary_df, use_index=True)
 
         salmon_table(tpm)
-        gene_txid, gene_name, geid_gname = get_translations(stringtie)
+
+        # Get translations for adding gene names to tables
+        (
+            txid_to_gene_name,  txid_to_gene_id, gid_to_gene_name
+         ) = get_translations(annotation)
 
         # Add gene names columns to counts files and write out
         # for publishing to user dir.
         df_dge = pd.read_csv(dge, sep='\t')
-        df_dge.insert(0, 'gene_name', df_dge.index.map(lambda x: geid_gname.get(x)))
+        df_dge.insert(0, 'gene_name', df_dge.index.map(
+            lambda x: gid_to_gene_name.get(x)))
         df_dge.to_csv('results_dge.tsv', index=True, index_label="gene_id", sep="\t")
 
-        # write_dge(gene_counts, geid_gname, "all_gene_counts.tsv")
+        # write_dge(gene_counts, gid_to_gene_name, "all_gene_counts.tsv")
         df_gene_counts = pd.read_csv(gene_counts, sep='\t')
         df_gene_counts.insert(
-                0, 'gene_name', df_gene_counts.index.map(lambda x: geid_gname.get(x)))
+            0, 'gene_name', df_gene_counts.index.map(
+                lambda x: gid_to_gene_name.get(x)))
         df_gene_counts.to_csv(
             'results_dge.tsv', index=True, index_label="gene_id", sep="\t")
 
         df_filtered = pd.read_csv(filtered, sep='\t')
         df_filtered.insert(1, "gene_name", df_filtered.gene_id.map(
-            lambda x: geid_gname.get(x)))
+            lambda x: gid_to_gene_name.get(x)))
         df_filtered.to_csv(
             'filtered_transcript_counts_with_genes.tsv', index=False, sep='\t')
 
         df_unfiltered = pd.read_csv(unfiltered, sep='\t')
         df_unfiltered.insert(1, "gene_name", df_unfiltered.gene_id.map(
-            lambda x: geid_gname.get(x)))
+            lambda x: gid_to_gene_name.get(x)))
         df_unfiltered.to_csv(
             'unfiltered_transcript_counts_with_genes.tsv', index=False, sep='\t')
 
         df_tpm = pd.read_csv(tpm, sep='\t')
         df_tpm.insert(1, "gene_name", df_tpm.Reference.map(
-            lambda x: gene_txid.get(x)))
+            lambda x: txid_to_gene_name.get(x)))
         df_tpm.to_csv("unfiltered_tpm_transcript_counts.tsv", index=False, sep='\t')
 
         # Add tables to report
-        dge_section(dge, gene_name, pval_threshold)
+        dge_section(df_dge, pval_threshold)
-        dexseq_section(dexseq, gene_name, pval_threshold)
+        dexseq_section(dexseq, txid_to_gene_name, txid_to_gene_id, pval_threshold)
-        dtu_section(dtu, gene_txid, gene_name)
+        dtu_section(dtu, txid_to_gene_name)

bin/workflow_glue/report.py

@@ -319,15 +319,16 @@ def de_section(report, de_report_dir, de_aln_stats_dir, pval_threshold):
     dexseq = de_report_dir / "results_dexseq.tsv"
     dge = de_report_dir / "results_dge.tsv"
     dtu = de_report_dir / "results_dtu_stageR.tsv"
-    # GFF file can have gtf or gff extension
+    # GFF file can have gtf or gff extension.
-    stringtie = next(de_report_dir.glob("*.g*f*"))
+    # Will be the original (transcriptome_source=precomputed) or wf-assembled annotation
+    annotation = next(de_report_dir.glob("*.g*f*"))
     tpm = de_report_dir / "unfiltered_tpm_transcript_counts.tsv"
     filtered = de_report_dir / "filtered_transcript_counts_with_genes.tsv"
     unfiltered = de_report_dir / "unfiltered_transcript_counts_with_genes.tsv"
     gene_counts = de_report_dir / "all_gene_counts.tsv"
     # This will also add a gene name column to the above counts tsv files
     de_plots.de_section(
-        stringtie=stringtie,
+        annotation=annotation,
         dexseq=dexseq,
         dge=dge,
         dtu=dtu,

bin/workflow_glue/tests/test_de_plots.py

@@ -0,0 +1,80 @@
+"""Test assign_barcodes."""
+from pathlib import Path
+
+import pytest
+from workflow_glue.de_plots import get_translations
+
+
+@pytest.fixture
+def test_data(request):
+    """Define data location fixture."""
+    return Path(request.config.getoption("--test_data")) / "workflow_glue"
+
+
+@pytest.mark.parametrize(
+    'annotation_file,expected',
+    [
+        [
+            'MSTRG.11088.gtf',
+            dict(gid_to_gene_name={
+                'ENSG00000236051.7': 'MYCBP2-AS1',
+                'ENSG00000283208.2': 'ENSG00000283208',
+                'ENSG00000102805.16': 'CLN5',
+                'MSTRG.11088': 'MSTRG.11088'
+            },
+                txid_to_gene_name={
+                    'ENST00000636183.2': 'CLN5',
+                    'ENST00000636780.2': 'CLN5',
+                    'ENST00000638147.2': 'ENSG00000283208',
+                    'ENST00000637192.1': 'ENSG00000283208',
+                    'ENST00000636737.1': 'MYCBP2-AS1',
+                    'ENST00000450627.6': 'MYCBP2-AS1',
+                    'MSTRG.11088.2': 'MSTRG.11088'
+                },
+                txid_to_gene_id={
+                    'ENST00000636183.2': 'ENSG00000102805.16',
+                    'MSTRG.11088.2': 'MSTRG.11088',
+                    'ENST00000636780.2': 'ENSG00000102805.16',
+                    'ENST00000638147.2': 'ENSG00000283208.2',
+                    'ENST00000637192.1': 'ENSG00000283208.2',
+                    'ENST00000636737.1': 'ENSG00000236051.7',
+                    'ENST00000450627.6': 'ENSG00000236051.7'
+                })
+
+        ],
+        # Small test to check that GFF3 works
+        [
+            'MSTRG.11088.gff3',
+            dict(gid_to_gene_name={
+                "ENSG00000290825.1": "DDX11L2",
+                "ENSG00000236397.3": "DDX11L2"
+            },
+                txid_to_gene_name={
+                    "ENST00000456328.2": "DDX11L2",
+                    "ENST00000437401.1": "DDX11L2"
+                },
+                txid_to_gene_id={
+                    'ENST00000437401.1': 'ENSG00000236397.3',
+                    'ENST00000456328.2': 'ENSG00000290825.1'
+                })
+        ]
+    ]
+)
+def test_get_translations(test_data, annotation_file, expected):
+    """Test that correct feature identifiers are extracted from the annotation.
+
+    `stringtie --merge` can sometimes generate gene models that may span multiple
+    reference genes. Possibly related issue:
+    https://github.com/gpertea/stringtie/issues/217
+    This can lead to the original genes and transcripts being assigned to that
+    incorrectly-merged gene model. The test data contains such a gene model generated
+    from `stringtie --merge` but actually consists of multiple different genes.
+
+
+    """
+    input_gtf = test_data / annotation_file
+    txid_to_gene_name, txid_to_gene_id, gid_to_gene_name = get_translations(input_gtf)
+
+    assert expected['gid_to_gene_name'] == gid_to_gene_name
+    assert expected['txid_to_gene_name'] == txid_to_gene_name
+    assert expected['txid_to_gene_id'] == txid_to_gene_id

nextflow.config

@@ -95,7 +95,7 @@ params {
       ]
     agent = null
     container_sha = "shad8671ea3a8ed52f2c0f40355e8eb5c6f00d2cbda"
-	common_sha="shaf15f9d80aba72c20e3e71f84869619873a56b8af"
+	common_sha="shabadd33adae761be6f2d59c6ecfb44b19cf472cfc"
     }
 }
 

subworkflows/differential_expression.nf

@@ -162,8 +162,9 @@ workflow differential_expression {
         analysis = deAnalysis(sample_sheet, merged, ref_annotation)
         plotResults(analysis.flt_counts, analysis.stageR, sample_sheet)
         // Concat files required for making the report
-        de_report = analysis.flt_counts.concat(analysis.gene_counts, analysis.dge, analysis.dexseq,
+        de_report = analysis.flt_counts.concat(
-        analysis.stageR, sample_sheet, merged, ref_annotation, merged_TPM, analysis.unflt_counts).collect()
+            analysis.gene_counts, analysis.dge, analysis.dexseq,
+            analysis.stageR, sample_sheet, merged, ref_annotation, merged_TPM, analysis.unflt_counts).collect()
         // Concat files required to be output to user without any changes
         de_outputs_concat = analysis.cpm.concat(plotResults.out.dtu_plots, analysis.dge_pdf, analysis.dge_tsv, analysis.dexseq,
         analysis.dtu_gene, analysis.dtu_transcript, analysis.dtu_stageR, analysis.dtu_pdf, analysis.flt_counts, analysis.gene_counts, merged_TPM).collect()

test_data/workflow_glue/MSTRG.11088.gff3

@@ -0,0 +1,2 @@
+chr1	HAVANA	transcript	11869	14409	.	+	.	ID=ENST00000456328.2;Parent=ENSG00000290825.1;gene_id=ENSG00000290825.1;transcript_id=ENST00000456328.2;gene_type=lncRNA;gene_name=DDX11L2;transcript_type=lncRNA;transcript_name=DDX11L2-202;level=2;transcript_support_level=1;tag=basic,Ensembl_canonical;havana_transcript=OTTHUMT00000362751.1
+chr2	HAVANA	transcript	113599036	113601261	.	-	.	ID=ENST00000437401.1;Parent=ENSG00000236397.3;gene_id=ENSG00000236397.3;transcript_id=ENST00000437401.1;gene_type=unprocessed_pseudogene;gene_name=DDX11L2;transcript_type=unprocessed_pseudogene;transcript_name=DDX11L2-201;level=2;transcript_support_level=NA;hgnc_id=HGNC:37103;ont=PGO:0000005;tag=basic,Ensembl_canonical;havana_gene=OTTHUMG00000047823.1;havana_transcript=OTTHUMT00000109036.1

test_data/workflow_glue/MSTRG.11088.gtf

@@ -0,0 +1,34 @@
+chr13	StringTie	transcript	76990660	77005117	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636183.2"; gene_name "CLN5"; ref_gene_id "ENSG00000102805.16"; 
+chr13	StringTie	exon	76990660	76992271	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636183.2"; exon_number "1"; gene_name "CLN5"; ref_gene_id "ENSG00000102805.16"; 
+chr13	StringTie	exon	76995063	76995228	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636183.2"; exon_number "2"; gene_name "CLN5"; ref_gene_id "ENSG00000102805.16"; 
+chr13	StringTie	exon	76995902	76996127	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636183.2"; exon_number "3"; gene_name "CLN5"; ref_gene_id "ENSG00000102805.16"; 
+chr13	StringTie	exon	77000458	77005117	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636183.2"; exon_number "4"; gene_name "CLN5"; ref_gene_id "ENSG00000102805.16";
+chr13	StringTie	transcript	76991729	77005117	1000	+	.	gene_id "MSTRG.11088"; transcript_id "MSTRG.11088.2"; 
+chr13	StringTie	exon	76991729	76991832	1000	+	.	gene_id "MSTRG.11088"; transcript_id "MSTRG.11088.2"; exon_number "1"; 
+chr13	StringTie	exon	76995063	76995228	1000	+	.	gene_id "MSTRG.11088"; transcript_id "MSTRG.11088.2"; exon_number "2"; 
+chr13	StringTie	exon	76995902	76996127	1000	+	.	gene_id "MSTRG.11088"; transcript_id "MSTRG.11088.2"; exon_number "3"; 
+chr13	StringTie	exon	77000458	77005117	1000	+	.	gene_id "MSTRG.11088"; transcript_id "MSTRG.11088.2"; exon_number "4";
+chr13	StringTie	transcript	76992044	77005117	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636780.2"; gene_name "CLN5"; ref_gene_id "ENSG00000102805.16"; 
+chr13	StringTie	exon	76992044	76992271	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636780.2"; exon_number "1"; gene_name "CLN5"; ref_gene_id "ENSG00000102805.16"; 
+chr13	StringTie	exon	76995063	76995228	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636780.2"; exon_number "2"; gene_name "CLN5"; ref_gene_id "ENSG00000102805.16"; 
+chr13	StringTie	exon	76995902	76996127	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636780.2"; exon_number "3"; gene_name "CLN5"; ref_gene_id "ENSG00000102805.16"; 
+chr13	StringTie	exon	76998043	76998085	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636780.2"; exon_number "4"; gene_name "CLN5"; ref_gene_id "ENSG00000102805.16"; 
+chr13	StringTie	exon	77000458	77005117	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636780.2"; exon_number "5"; gene_name "CLN5"; ref_gene_id "ENSG00000102805.16";
+chr13	StringTie	transcript	76992078	77078025	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000638147.2"; gene_name "ENSG00000283208"; ref_gene_id "ENSG00000283208.2";
+chr13	StringTie	exon	76992078	76992271	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000638147.2"; exon_number "1"; gene_name "ENSG00000283208"; ref_gene_id "ENSG00000283208.2"; 
+chr13	StringTie	exon	76995063	76995228	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000638147.2"; exon_number "2"; gene_name "ENSG00000283208"; ref_gene_id "ENSG00000283208.2"; 
+chr13	StringTie	exon	76995902	76996127	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000638147.2"; exon_number "3"; gene_name "ENSG00000283208"; ref_gene_id "ENSG00000283208.2"; 
+chr13	StringTie	exon	77075518	77075584	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000638147.2"; exon_number "4"; gene_name "ENSG00000283208"; ref_gene_id "ENSG00000283208.2"; 
+chr13	StringTie	exon	77076816	77078025	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000638147.2"; exon_number "5"; gene_name "ENSG00000283208"; ref_gene_id "ENSG00000283208.2";
+chr13	StringTie	transcript	76995915	77129717	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000637192.1"; gene_name "ENSG00000283208"; ref_gene_id "ENSG00000283208.2"; 
+chr13	StringTie	exon	76995915	76996127	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000637192.1"; exon_number "1"; gene_name "ENSG00000283208"; ref_gene_id "ENSG00000283208.2"; 
+chr13	StringTie	exon	77109648	77110102	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000637192.1"; exon_number "2"; gene_name "ENSG00000283208"; ref_gene_id "ENSG00000283208.2"; 
+chr13	StringTie	exon	77129147	77129717	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000637192.1"; exon_number "3"; gene_name "ENSG00000283208"; ref_gene_id "ENSG00000283208.2";
+chr13	StringTie	transcript	77026767	77078025	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636737.1"; gene_name "MYCBP2-AS1"; ref_gene_id "ENSG00000236051.7";
+chr13	StringTie	exon	77026767	77027122	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636737.1"; exon_number "1"; gene_name "MYCBP2-AS1"; ref_gene_id "ENSG00000236051.7"; 
+chr13	StringTie	exon	77075518	77075584	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636737.1"; exon_number "2"; gene_name "MYCBP2-AS1"; ref_gene_id "ENSG00000236051.7"; 
+chr13	StringTie	exon	77076816	77078025	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000636737.1"; exon_number "3"; gene_name "MYCBP2-AS1"; ref_gene_id "ENSG00000236051.7"; 
+chr13	StringTie	transcript	77075514	77087778	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000450627.6"; gene_name "MYCBP2-AS1"; ref_gene_id "ENSG00000236051.7"; 
+chr13	StringTie	exon	77075514	77075584	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000450627.6"; exon_number "1"; gene_name "MYCBP2-AS1"; ref_gene_id "ENSG00000236051.7"; 
+chr13	StringTie	exon	77076816	77076866	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000450627.6"; exon_number "2"; gene_name "MYCBP2-AS1"; ref_gene_id "ENSG00000236051.7"; 
+chr13	StringTie	exon	77087552	77087778	1000	+	.	gene_id "MSTRG.11088"; transcript_id "ENST00000450627.6"; exon_number "3"; gene_name "MYCBP2-AS1"; ref_gene_id "ENSG00000236051.7";