IndexError when running spladder

[swu13]

• spladder version:3.0.1
• Python version:3.6
• Operating System:CentOS

Description

I use spladder to detect splicing events in one sample with bam file, The command is: spladder build --parallel 20 -b *bam -a gencode.v32.annotation.gtf -o .

But it shows index error below. And one of the file (merge_graphs_mult_exon_skip_C3.confirmed.txt.gz) has incomplete information. The last line is only "chr2 + mult_exon_skip.4684 3 ENSG00000064012.21" and contains no alternative splicing events.

What I Did

command(s) 
1. STAR --genomeDir ./STAR_index/  --readFilesIn XX.R1.fastq XX.R2.fastq --runThreadN 20 --outFilterMultimapScoreRange 1 --outFilterMultimapNmax 20 --outFilterMismatchNmax 10 --alignIntronMax 500000 --alignMatesGapMax 1000000 --sjdbScore 2 --alignSJDBoverhangMin 1 --genomeLoad NoSharedMemory --limitBAMsortRAM 70000000000 --readFilesCommand cat --outFilterMatchNminOverLread 0.33 --outFilterScoreMinOverLread 0.33 --sjdbOverhang 100 --outSAMstrandField intronMotif --outSAMattributes NH HI NM MD AS XS --sjdbGTFfile gencode.v32.annotation.gtf --limitSjdbInsertNsj 2000000 --outSAMunmapped None --outSAMtype BAM SortedByCoordinate --outSAMheaderHD @HD VN:1.4  --twopassMode Basic --outSAMmultNmax 1   --quantMode TranscriptomeSAM   --outFileNamePrefix .
2. spladder build --parallel 20 -b *bam -a gencode.v32.annotation.gtf -o .



If there was a crash, please include the traceback here.
``Reporting confirmed mult_exon_skip events:
writing mult_exon_skip events in gff3 format to .//merge_graphs_mult_exon_skip_C3.confirmed.gff3
writing mult_exon_skip events in flat txt format to .//merge_graphs_mult_exon_skip_C3.confirmed.txt.gz
Traceback (most recent call last):
  File "/data/myang9/software/miniconda3/bin/spladder", line 8, in <module>
    sys.exit(main())
  File "/data/myang9/software/miniconda3/lib/python3.9/site-packages/spladder/spladder.py", line 229, in main
    options.func(options)
  File "/data/myang9/software/miniconda3/lib/python3.9/site-packages/spladder/spladder_build.py", line 163, in spladder
    analyze_events(event_type, options.bam_fnames, options)
  File "/data/myang9/software/miniconda3/lib/python3.9/site-packages/spladder/alt_splice/analyze.py", line 191, in analyze_events
    write_events_txt(fn_out_conf_txt, options.samples[sample_idx], events_all, fn_out_count, event_idx=confirmed_idx)
  File "/data/myang9/software/miniconda3/lib/python3.9/site-packages/spladder/alt_splice/write.py", line 89, in write_events_txt
    counts = event_counts_chunk[:, :, i - chunk_idx_event[0]]
IndexError: index 1344 is out of bounds for axis 2 with size 1113`

[lunazhaoxxx]

I have met the same error using command: spladder build -b xxx.bam -a gencode.v39.annotation.gtf -o xxx
which says:
Reporting confirmed exon_skip events:
writing exon_skip events in gff3 format to xxx/merge_graphs_exon_skip_C3.confirmed.gff3
writing exon_skip events in flat txt format to xxx/merge_graphs_exon_skip_C3.confirmed.txt.gz
Traceback (most recent call last):
File "/usr/local/bin/spladder", line 11, in
load_entry_point('spladder==3.0.2', 'console_scripts', 'spladder')()
File "/usr/local/lib/python3.8/dist-packages/spladder-3.0.2-py3.8.egg/spladder/spladder.py", line 229, in main
options.func(options)
File "/usr/local/lib/python3.8/dist-packages/spladder-3.0.2-py3.8.egg/spladder/spladder_build.py", line 163, in spladder
analyze_events(event_type, options.bam_fnames, options)
File "/usr/local/lib/python3.8/dist-packages/spladder-3.0.2-py3.8.egg/spladder/alt_splice/analyze.py", line 191, in analyze_events
write_events_txt(fn_out_conf_txt, options.samples[sample_idx], events_all, fn_out_count, event_idx=confirmed_idx)
File "/usr/local/lib/python3.8/dist-packages/spladder-3.0.2-py3.8.egg/spladder/alt_splice/write.py", line 90, in write_events_txt
psi = psi_chunk[:, i - chunk_idx_psi[0]]
IndexError: index 2354 is out of bounds for axis 1 with size 1015

[huguesfontenelle]

I'm running into this as well, unfortunately.
I placed a breakpoint() around the

counts = event_counts_chunk[:, :, i - chunk_idx_event[0]]

line, but can't really figure out how it works.

@akahles @warrenmcg @izcram @ratsch
What would you need to reproduce this? Would the following files be sufficient:

• merge_graphs_exon_skip_C3.confirmed.pickle
• merge_graphs_exon_skip_C3.counts.hdf5
• merge_graphs_exon_skip_C3.pickle

Thanks!

[dlsoltero]

I got this same problem while testing a sample of a bam, but when I tried the complete bam it worked ok. I just mention it because it may help to figure out the problem.

[riasc]

Has anyone resolved this? I have it both on small datasets and the whole BAM file. I discovered that the problem occurs mainly when aligning with STAR, but not when using BWA. Is there something that needs to be considered?

It occurs on exon_skip events:

Reporting confirmed exon_skip events:
writing exon_skip events in gff3 format to testspladderSTAR_exonskip/merge_graphs_exon_skip_C3.confirmed.gff3
writing exon_skip events in flat txt format to testspladderSTAR_exonskip/merge_graphs_exon_skip_C3.confirmed.txt.gz
Traceback (most recent call last):
File "/home/sej9799/mambaforge/envs/spladder/bin/spladder", line 8, in
sys.exit(main())
File "/home/sej9799/mambaforge/envs/spladder/lib/python3.10/site-packages/spladder/spladder.py", line 229, in main
options.func(options)
File "/home/sej9799/mambaforge/envs/spladder/lib/python3.10/site-packages/spladder/spladder_build.py", line 163, in spladder
analyze_events(event_type, options.bam_fnames, options)
File "/home/sej9799/mambaforge/envs/spladder/lib/python3.10/site-packages/spladder/alt_splice/analyze.py", line 191, in analyze_events
write_events_txt(fn_out_conf_txt, options.samples[sample_idx], events_all, fn_out_count, event_idx=confirmed_idx)
File "/home/sej9799/mambaforge/envs/spladder/lib/python3.10/site-packages/spladder/alt_splice/write.py", line 89, in write_events_txt
counts = event_counts_chunk[:, :, i - chunk_idx_event[0]]
IndexError: index 7879 is out of bounds for axis 2 with size 2545

[huguesfontenelle]

I solved that for "my" case.

Context: I use a BAM file with a single chromosome for testing.
Possible cause: The GTF annotation file refers chromosomes / contigs that are not present in the BAM file.
Solution: slice the GTF to include only the chromosomes / contigs present in the BAM.

[akahles]

Thanks a lot for reporting this. I have difficulty reproducing the issue. Would one of you be able to provide a minimal failing example? This would greatly speed up the debugging on my end. Usually, it should not be a problem if the chromosome/contig sets in alignment and annotation files are not the same. As long as the intersection is not empty, SplAdder should generate output.

Best,
Andre

[riasc]

So this basiclally means that no output can be generated? Is there a way to generate normal normal even when no output is going to be generated?