How to quantify the RNA-seq reads using Kallisto?

Question

[Farah]

Hello, I am running Kallisto (for three paired files) as below. However, I got the following Errors:

mkdir /scratch/fs/ye/kallisto
cd /scratch/fs/ye/kallisto

find /scratch/fs/ye/trimmomatic -name "*trimmomatic_[12].fq.gz" | sort | head -n 6 | while read FW_READ
do
 read RV_READ
 FILEBASE=$(basename "${FW_READ%_1.fq.gz}")
 kallisto quant -i /home/fs/Homo_sapiens.GRCh38.cdna.all.release-97.idx -b 100 \
 o . -t 16 "$FW_READ" "$RV_READ"

 # because Kallisto doesn't let us specify an output filename so we rename all
 # output files

 mv "abundance.tsv" $FILEBASE-"abundance.tsv"

 mv "abundance.h5" $FILEBASE-"abundance.h5"

 mv "run_info.json" $FILEBASE-"run_info.json"

done

The ERRORs are as follows:

Error: file not found o
Error: need to specify output directory

mv: cannot stat `abundance.tsv': No such file or directory

mv: cannot stat `abundance.h5': No such file or directory

mv: cannot stat `run_info.json': No such file or directory

I would highly appreciate if you can help me how to fix this error.

Thank you very much.

Answer 1

[Ram]

Does your kallisto command need a -o . instead of o .? That's what the error message seems to be saying.

Plus, why the read RV_READ? Can that not be derived from FW_READ?

[Farah]

Dear RamRS, Thank you so much for your advice. As you suggested, I used -o . instead of o . It generated the below messages in my standard error output file:

[quant] fragment length distribution will be estimated from the data
[index] k-mer length: 31
[index] number of targets: 189,154
[index] number of k-mers: 109,813,275
[index] number of equivalence classes: 763,796
[quant] running in paired-end mode
[quant] will process pair 1: /scratch/fs/ye/trimmomatic/SRR6822797-sortmerna-trimmomatic_1.fq.gz
                             /scratch/fs/ye/trimmomatic/SRR6822797-sortmerna-trimmomatic_2.fq.gz
[quant] finding pseudoalignments for the reads ... done
[quant] processed 12,755,479 reads, 11,205,135 reads pseudoaligned
[quant] estimated average fragment length: 161.419
[   em] quantifying the abundances ... done
[   em] the Expectation-Maximization algorithm ran for 1,258 rounds
[bstrp] number of EM bootstraps complete: 100

[quant] fragment length distribution will be estimated from the data
[index] k-mer length: 31
[index] number of targets: 189,154
[index] number of k-mers: 109,813,275
[index] number of equivalence classes: 763,796
[quant] running in paired-end mode
[quant] will process pair 1: /scratch/fs/ye/trimmomatic/SRR6822798-sortmerna-trimmomatic_1.fq.gz
                             /scratch/fs/ye/trimmomatic/SRR6822798-sortmerna-trimmomatic_2.fq.gz
[quant] finding pseudoalignments for the reads ... done
[quant] processed 5,362,940 reads, 4,798,586 reads pseudoaligned
[quant] estimated average fragment length: 166.668
[   em] quantifying the abundances ... done
[   em] the Expectation-Maximization algorithm ran for 1,057 rounds
[bstrp] number of EM bootstraps complete: 100


[quant] fragment length distribution will be estimated from the data
[index] k-mer length: 31
[index] number of targets: 189,154
[index] number of k-mers: 109,813,275
[index] number of equivalence classes: 763,796
[quant] running in paired-end mode
[quant] will process pair 1: /scratch/fs/ye/trimmomatic/SRR6822799-sortmerna-trimmomatic_1.fq.gz
                             /scratch/fs/ye/trimmomatic/SRR6822799-sortmerna-trimmomatic_2.fq.gz
[quant] finding pseudoalignments for the reads ... done
[quant] processed 4,315,257 reads, 3,750,150 reads pseudoaligned
[quant] estimated average fragment length: 164.278
[   em] quantifying the abundances ... done
[   em] the Expectation-Maximization algorithm ran for 1,242 rounds
[bstrp] number of EM bootstraps complete: 100

It also produced abundance.tsv, abundance.h5, and run_info.json files for each of my paired files. But, how can I be sure that all of these files are correct, and how to check if kallisto was performed correctly? as my standard error output is not zero, and it has a size of 15.

Many thanks for your help.

[Ram]

STDERR is not a placeholder for just error messages. It logs everything that does not belong in STDOUT. Progress messages, warnings and error messages are examples of what could go into STDERR. Do not judge your program status by STDERR file size.

I am not familiar with kallisto but if the above is all of your STDERR, the program seems to have completed OK (I don't see any error message).

[Farah]

Many thanks for your help. Yes you are right, it is now working well. Thanks a lot.