I am contacted you regarding some questions of "mrfQuantifier" in RSEQtools.
We have cloned many novel transcripts from the whole human brain and wanted
to relate their expression to specific periods/regions using RNA-seq data
from the Brainspan. We were able to download the MRF file that only contains
alignment blocks from Brainspan. We have noticed that you have described in
the Bioinformatics paper that mrfQuantifier does not perform the expression
quantification on the transcript level.
So, Could you let me know why mrfQuantifier does not perform on the
transcript level? Do you mean it does not address the problem of multireads?
Also, "mrf2sam" does not work with MRF files we downloaded from Brainspain
and it shows error "PROBLEM: Unknown presentColumn:
H376_V_51_A1C_L_RNASeq.mrf". Could that be it needs the sequences
information? Also, is any recent tool in RESQtools that would perform on the
transcript level for expression quantification?
iqseq can perform transcript level quant. Here is my understanding to your questions.
1) When RSEQtools were first published in 2011, it was designed to just take care of the gene or exon form level quantification. Isoform quantification is challenging, partially because of the multi-reads.
Then in 2012, our lab published another paper IQseq (https://code.google.com/p/iqseq/downloads/list). IQseq actually take all the reads mapped to one gene, for those mapped to multiple isoforms, EM algorithm is used to allocate the reads iteratively to each transcript.
2) For the second question, I am not sure what is happening. If possible could please send me the sample mrf files for me to check what’s going on there?
Hopefully this could help a little bit. Please let me know if you have any questions.
We have our novel transcript annotations (eg, exon boundaries), and we downloaded all MRF format files of RNA-seq reads of developmental brain from Brainspan to try to quantify the expression levels.
Under "Supplemental Data" from http://www.brainspan.org/static/download.html , there are all the MRF files we downloaded. I have tried to run them with mrf2sam and failed with error.
My question is, would IQseq tool works with MRF files containing only alignment blocks (genomic coordinates only) without sequence information since Brainspan MRF files contain no such information? or it still needs sequences and quality score information to be able to run the tool. The complete set of MRF files can be obtained from Under "Supplemental Data" from http://www.brainspan.org/static/download.html for testing. If would be great if you could also check whether or not IQSeq would work on these set of MRF files.
After consulting some of my colleagues, I have the answer like this.
For one of the files, if you type
mrf2sam <H376_IIA_50_LGE_L_RNASeq.mrf H376_IIA_50_LGE_L_RNASeq.sam
it works, but if you type mrf2sam H376_IIA_50_LGE_L_RNASeq.mrf
it wont work.
I am not the author of RSEQtools, but I guess that the reason for the command line to be like this is that everything should be piped to avoid I/O for the middle files.
If you still can not figure it out, please let me know what it your command and the error message. Hope this would help!