rsem + star pipeline in yeast (S cerevisiae) samples

[Marco Di Stefano]

Hi all,

I'm analysing a series of yeast samples. I have the following problem in the quantification step.

1) To generate the index I used 

 

STAR --runMode genomeGenerate --runThreadN 8 --genomeDir . --genomeFastaFiles XXX.fa --sjdbGTFfile XXX.gtf --outFileNamePrefix XXXprefixXXX --genomeSAindexNbases 10

I noticed that I'm not able to use an analogous rsem-prepare-reference command because I need to change the default value for --genomeSAindexNbases from 14 to 10.

2) Map with STAR.

3) To quantify using 

rsem-calculate-expression -p 8 --no-bam-output --bam  --paired-end XXXbamfileXXX XXXrefnameXXX XXXsamplenameXXX 

rsem is looking for the *grp file which has not been generated at point 1 so it doesn't exist.

Since I really need to change the option --genomeSAindexNbases at step 1, is there a way to pass this option to rsem-prepare-reference?

Thanks,

Marco

[Bo Li]

Hi Marco,

Before you run rsem-calculate-expression, you have to run
rsem-prepare-reference. However, you do not need to use RSEM to
build/run STAR aligner for you. You just need to make sure the extracted
reference transcripts using rsem-prepare-reference and STAR are the
same.

Hope it helps,
Bo

> --
> RSEM website: http://deweylab.biostat.wisc.edu/rsem/ [1]
> ---
> You received this message because you are subscribed to the Google
> Groups "RSEM Users" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to rsem-users+...@googlegroups.com.
> To post to this group, send email to rsem-...@googlegroups.com.
> Visit this group at https://groups.google.com/group/rsem-users [2].
>
>
> Links:
> ------
> [1] http://deweylab.biostat.wisc.edu/rsem/
> [2] https://groups.google.com/group/rsem-users

[Marco Di Stefano]

Hi Bo,

Thanks for your reply.

To implement your suggestion, I compared the files transcriptInfo.tab generated using

1 - STAR --runMode genomeGenerate --runThreadN 8 --genomeDir . --genomeFastaFiles XXX.fa --sjdbGTFfile XXX.gtf --outFileNamePrefix XXXprefix --genomeSAindexNbases 10

and

2 - rsem-prepare-reference --star -p 8 --gtf XXX.gtf XXX.fa XXX

Since they are the same, I can safely use the *grp file generated by 2 even if the mapping has been done using the Index generated by 1.

Is this ok?

Excuse me for being pedantic and thanks a lot,

Marco

[Bo Li]

Hi Marco,

What's this transcriptInfo.tab file?

Thanks,
Bo

[Marco Di Stefano]

Hi Bo,

The transcriptInfo.tab is part of the output of STAR genome generation. It reports on some features of the 

transcripts mapped onto the reference genome. 

E.g. These are the first 3 lines of one of those files:

7126

YAL069W         334 648 648 1 1 0

YAL068W-A 537 791 648 1 1 1

YAL068C         1806 2168 791 2 1 2

I thought that, since this file is the same using rsem-prepare-reference or STAR, I could conclude that the extracted 

reference transcripts are the same.

If this is not the case, what can I do to check whether the extracted reference transcripts using rsem-prepare-reference 

and STAR are the same? Thanks a lot for your help.

Cheers,

Marco

[Bo Li]

Hi Marco,

This is not the case. You want to make sure the GTF file provided to the
STAR aligner is the same as the GTF file provided to the
'rsem-prepare-reference'. In particular, this is the STAR commands RSEM
uses to build STAR indices:

$command = $star_path . "STAR " .
" --runThreadN $star_nthreads " .
" --runMode genomeGenerate " .
" --genomeDir $out_star_genome_path " .
" --genomeFastaFiles @list " .
" --sjdbGTFfile $gtfF " .
" --sjdbOverhang $star_sjdboverhang " .
" --outFileNamePrefix $ARGV[1]";

Hope it helps,
Bo

[Marco Di Stefano]

Hi Bo,

This clarified me what I needed for my analysis.

Thanks a lot,

Marco