rsem-run-ebseq Error in min(EachChoose0[which(EachChoose0 != -Inf)])

[Pankaj Agarwal]

Hi,

I am trying to use rsem-run-ebseq for isoform differential expression analysis.  

Using the rsem-generate-data-matrix the count matrix file was generated as:

rsem-generate-data-matrix \

./isoform_level/SRR1617647.isoforms.results ./isoform_level/SRR1617647.isoforms.results ./isoform_level/SRR1617647.isoforms.results ./isoform_level/SRR1617647.isoforms.results ./isoform_level/SRR1617647.isoforms.results ./isoform_level/SRR1617647.isoforms.results ./isoform_level/SRR1617647.isoforms.results ./isoform_level/SRR1617647.isoforms.results ./isoform_level/SRR1617647.isoforms.results \

> SRR.counts.matrix  

Using rsem-generate-ngvector the ngvector was generates as:

rsem-generate-ngvector hg19.transcripts.fa hg19 

which generated two files:

-rw-r--r-- 1 agarw005 agarw005 96018 Mar 31 11:09 hg19.ngvec

-rw-r--r-- 1 agarw005 agarw005 857904 Mar 31 11:09 hg19.ump

Then rsem-run-ebseq was run as follows:

rsem-run-ebseq --ngvector hg19.ngvec SRR.counts.matrix 6,3 SRR_iso_test_033115A.results

rsem-for-ebseq-find-DE /home/agarw005/rsem/rsem-1.2.19/EBSeq hg19.ngvec SRR.counts.matrix SRR_iso_test_033115A.results 6 3

Loading required package: blockmodeling

Loading required package: gplots

Attaching package: ‘gplots’

The following object is masked from ‘package:stats’:

lowess

Removing transcripts with 75 th quantile < = 10

29520 transcripts will be tested

Error in min(EachChoose0[which(EachChoose0 != -Inf)]) :

invalid 'type' (list) of argument

Calls: EBTest

Execution halted

"rsem-for-ebseq-find-DE /home/agarw005/rsem/rsem-1.2.19/EBSeq hg19.ngvec SRR.counts.matrix SRR_iso_test_033115A.results 6 3" failed! Plase check if you provide correct parameters/options for the pipeline!

I cannot find anything related to this error on the forum or google.

In case it helps, here is a sample from the matrix count file:

"isoform_level/SRR1617647.isoforms.results" "isoform_level/SRR1617647.isoforms.results" "isoform_level/SRR1617647.isoforms.results" "isoform_level/SRR1617647.isoforms.results" "isoform_level/SRR1617647.isoforms.results" "isoform_level/SRR1617647.isoforms.results" "isoform_level/SRR1617647.isoforms.results" "isoform_level/SRR1617647.isoforms.results" "isoform_level/SRR1617647.isoforms.results"

"NR_001526" 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

"NR_001526_1" 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

"NR_001526_2" 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

"NM_130786" 659.00 659.00 659.00 659.00 659.00 659.00 659.00 659.00 659.00

"NR_015380" 816.50 816.50 816.50 816.50 816.50 816.50 816.50 816.50 816.50

"NM_001198818" 32.00 32.00 32.00 32.00 32.00 32.00 32.00 32.00 32.00

"NM_001198819" 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

"NM_001198820" 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

"NM_014576" 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

"NM_138932" 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

"NM_138933" 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

"NM_000014" 27378.00 27378.00 27378.00 27378.00 27378.00 27378.00 27378.00 27378.00 27378.00

"NR_026971" 141.00 141.00 141.00 141.00 141.00 141.00 141.00 141.00 141.00

"NM_001282424" 45.02 45.02 45.02 45.02 45.02 45.02 45.02 45.02 45.02

...

...

Should I just use the EBSeq package in R?  I am wondering if I will have the same problem.

Thank you for any help or suggestion you can provide.

- Pankaj

[Ning Leng]

Hi Pankaj,

I think the error might indicate that all genes (isoforms ) in your data have variance 0. Seems the data set is multiple repeats of the same sample?

Best,

Ning

--
RSEM website: http://deweylab.biostat.wisc.edu/rsem/
---
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 http://groups.google.com/group/rsem-users.

[Pankaj Agarwal]

Hi Ning,

Yes, you are right.  I made a mistake when generating the count matrix and put the same sample name for all.  Now I fixed it and it ran fine.

Could you please help me interpret the results in the output file.  The first few lines are:

"PPEE"  "PPDE"  "PostFC"        "RealFC"

"NM_001164750"  0       1       0.00119265289397819     0.000397066870965699

"NM_003836"     0       1       0.00165959521854055     0.000938642190345861

"NM_002281"     0       1       0.0113815352350426      0.00634097567795835

"NM_002457"     0       1       0.00611924005645253     0.00298318803679357

"NM_058173"     0       1       0.00488345050932763     0.00287141908900717

"NM_144686"     0       1       2.53252114465571        2.5321300433228

"NM_013387"     0       1       0.259552428071597       0.258004796637875

"NM_001127505"  2.22044604925031e-16    1       0.00215632071358735     0.000604027656497426

"NM_213651"     1.55431223447522e-15    0.999999999999998       0.00200908900001523     0.000743447504594512

"NM_001144945"  2.33146835171283e-15    0.999999999999998       0.00272387016121283     0.00167964291609369

"NM_002985"     3.44169137633799e-15    0.999999999999997       0.256276003692597       0.252545126116836

"NM_001079522"  5.21804821573824e-15    0.999999999999995       0.000434312017838965    0.000143711042216573

...

...

Could you please let me know the interpretation of column headings

"PPEE"  "PPDE"  "PostFC"        "RealFC"

and is multiple correction taken into account?

Thank you,

- Pankaj