RSEM calculate expression troubleshoot

Question

[prabin.dm]

Hi, I am trying to use RSEM following this [tutorial]

This is the error code I am getting.

readline() on closed filehandle INPUT at /share/pkg/RSEM/1.3.0/bin/rsem-calculate-expression line 350.
Use of uninitialized value $line in scalar chomp at /share/pkg/RSEM/1.3.0/bin/rsem-calculate-expression line 350.
Use of uninitialized value $line in split at /share/pkg/RSEM/1.3.0/bin/rsem-calculate-expression line 352.
Use of uninitialized value $type in numeric ne (!=) at /share/pkg/RSEM/1.3.0/bin/rsem-calculate-expression line 353.

Can anyone explain what this means?

Here is the command I used. I should also mention that I have not used the same datasets in the tutorial I downloaded a different dataset and trying to use it. dataset1

The fastq files were QC-ed with FastQC and trimmed with trimmomatric

# LSBATCH: User input
#!/bin/bash

#BSUB -q long
#BSUB -W 2:0 
#BSUB -n 8 
#BSUB -R "rusage[mem=2048] span[ptile=2]"
#BSUB -J rsem_SRR836120
#BSUB -o out.%J
#BSUB -e err.%J

module load RSEM/1.3.0
module load bowtie2/2.3.2

rsem-calculate-expression -p 8  \
                --bowtie2 --bowtie2-path /share/pkg/bowtie2/2.3.2 \
                --estimate-rspd \
                --append-names \
                --output-genome-bam \
                SRR836120_1.fastq_trim.fastq \
               /ref/mouse_ref \
              /SRR836120_Rsem

module unload RSEM/1.3.0
module unload bowtie2/2.3.2

[Nitin Narwade]

Snippet from rsem-calculate-expression perl scipt.

348 if ($genGenomeBamF) {
349 open(INPUT, "$refName.ti");
350 my $line = <INPUT>; chomp($line);
351 close(INPUT);
352 my ($M, $type) = split(/ /, $line);
353 pod2usage(-msg => "No genome information provided, so genome bam file cannot be generated!\n", -exitval => 2, -verbose => 2) if ($type != 0);
354 }

On the line 350 of rsem-calculate-expression script is indicating that reading some content from INPUT file handler.

The usual error readline() on closed filehandle is raise because PERL is unable to open the input file and it subsequently fails to read the contents from it.

No genome information provided, so genome bam file cannot be generated, I think there is some problem with your reference file.

Does your program interrupt?

[prabin.dm]

Thanks Nitin,

I went back and checked there genome reference. And used a prebuild reference. And it worked fine.

[h.mon]

What is the command-line you used? What tutorial are you following, the link didn't get posted.

[prabin.dm]

I updated the post with details. Thank you for your response.

[igor]

Did you try running the command interactively (without submitting it as a job)?

[prabin.dm]

Was able to solve it by using a prebuilt reference. Thanks for your response

[bjwiley23]

Did you happen to build the reference genome index with the second option: building references from a set of transcripts? I get your same same error No genome information provided but only when I build using option 2.

[bjwiley23]

Wow :) I just read down further. This explains why I get the error only with option 2 building reference from a set of transcripts:

"Thus, we turn on the --output-genome-bam option in the fifth line. Note that this option is only available when you build references from a genome."

Therefore when building from set of transcripts do not use this argument.