Error in `$<-.data.frame`(`*tmp*`, "filterInterpretation", value = character(0))

[maressyl]

Hi,

I'm trying to run the latest version of SpliceLauncher (93ec1c5) but I face an R error during the last step (full stdout and stderr log attached) :

270650.stdout.txt
270650.stderr.txt

######################
#Statistical analysis...
######################
   Creating matrix...
   Fitting model...
   Model apliccation...
   tmp file with p-values...
   Following modeled junctions are not in the models: 0 %
Error in `$<-.data.frame`(`*tmp*`, "filterInterpretation", value = character(0)) : 
  replacement has 0 rows, data has 130
Calls: $<- -> $<-.data.frame
Execution halted

Alignment was performed on each sample individually by SpliceLauncher :

image=".../SpliceLauncher-93ec1c5/splicelauncher.sif"
script=".../SpliceLauncher-93ec1c5/SpliceLauncher.sh"
config=".../SpliceLauncher-93ec1c5/config.cfg"

ulimit -S -n 10000
singularity exec --bind /srv "${image}" bash "${script}" \\
   --runMode Align \\
   -C "${config}" \\
   -F "${sampleDir}" \\
   -O "${sortie}/$(basename "$sampleDir")" \\
   -p -t 5

And SpliceLauncher was run with :

singularity exec --bind /srv ".../SpliceLauncher-93ec1c5/splicelauncher.sif" bash ".../SpliceLauncher-93ec1c5/SpliceLauncher.sh" \
	--runMode Count,SpliceLauncher \
	-C ".../SpliceLauncher-93ec1c5/config.cfg" \
	-B "$bamDir" \
	-O "$sortie" \
	-p --Graphics --removeOther

Singularity image was built yesterday from the provided recipe, and installation was performed using the following files :

• https://ftp.ncbi.nlm.nih.gov/refseq/H_sapiens/annotation/GRCh38_latest/refseq_identifiers/GRCh38_latest_genomic.gff.gz
• https://ftp.ncbi.nlm.nih.gov/refseq/H_sapiens/annotation/GRCh38_latest/refseq_identifiers/GRCh38_latest_genomic.fna.gz
• https://ftp.ncbi.nlm.nih.gov/refseq/MANE/MANE_human/current/MANE.GRCh38.v1.4.summary.txt.gz
• https://ftp.ncbi.nlm.nih.gov/refseq/H_sapiens/annotation/GRCh38_latest/refseq_identifiers/GRCh38_latest_assembly_report.txt

singularity exec --bind /srv splicelauncher.sif bash ./SpliceLauncher.sh \
   --runMode INSTALL \
   -O ./GRCh38-2025 \
   --samtools samtools --bedtools bedtools --STAR STAR \
   --fasta ../GRCh38_latest_genomic.sub.fna \
   --gff ../GRCh38_latest_genomic.sub.gff \
   --threads 30 \
   --assembly_report ../GRCh38_latest_assembly_report.txt \
   --mane ../MANE.GRCh38.v1.4.summary.sub.txt

FASTQ files were produced by simulation and can be shared if necessary (30 samples of 16 million read pairs).

Similar files went fine with an older version of SpliceLauncher (commit d0486847c3dc60064864f7fa528638e3ab1c68c5) and annotation (september 2022) but I wanted to try latest improvements.

Is there something you can do or suggest ?

Best regards,
Sylvain

[maressyl]

I tried to dig a bit more, a few observations :

• output/output.txt is surprisingly small (35,637 rows) as compared to the one produced on same data by previous SpliceLauncher version (207,212 rows)
• BEDannotation.bed is surprisingly small (42,065 rows) as compared to the one generated with previous SpliceLauncher version (2,083,520 rows).