A problem with MMSEQ

[Leo.Zhang]

I use mmseq package, and happen the follow problem

[leo@liu mmseq]$ head SR016359.1.exp.gene.mmseq
# Mapped reads (or read pairs): 6120659
gene_id log_mu_em log_mu_gibbs mcse weighted_length
ENSG00000000003 1.23277 1.22626 0.00597861 1784.14
ENSG00000000005 NA NA NA NA
ENSG00000000419 2.22127 2.11945 0.00583188 1040.51
ENSG00000000457 0.958256 0.837237 0.00850429 2854.55
ENSG00000000460 0.657037 0.615422 0.00951234 3696.15
ENSG00000000938 1.68165 1.59352 0.00655677 1745.39
ENSG00000000971 1.23488 1.23017 0.00478502 4283.87
ENSG00000001036 1.97813 1.8715 0.00458929 1654.84


[leo@liu mmseq]$ head SR016359.1.exp.mmseq
# Mapped reads (or read pairs): 6120659
transcript_id log_mu_em log_mu_gibbs mcse length
ENST00000539570 -inf -11.5734 0.33027 710
ENST00000576455 -0.890898 -6.78479 0.355641 2012
ENST00000510508 -180.836 -11.0123 0.322965 881
ENST00000474471 0.71319 -6.78531 0.40447 1175
ENST00000381700 NA NA NA 320
ENST00000445946 -1.1245 -1.5426 0.03394 503
ENST00000472572 -40.0731 -7.81427 0.39148 1052
ENST00000420022 0.828722 0.742549 0.0130477 428


what is meaning for "NA" in output file?

My command line:
bowtie -f -a -S -m 100 --fullref -p 2 /home/leo/BitSeq/Homo_sapiens.GRCh37.67.cdna.all \
/home/leo/cufflinks/MAQC.RawData/SRX016359.1.fasta | samtools view -F 0xC -bS - | samtools sort -n - SR016359.12.namesorted

bam2hits -m "(E\S+).*geneE\S+).*" 1 2 Homo_sapiens.GRCh37.67.cdna.all.fa SR016359.12.namesorted.bam > SR016359.12.hits

mmseq SR016359.12.hits SR016359.1

[eturro]

"NA" means the feature (in your case, the Ensembl transcript) had zero reads aligned to it, which means its true expression is likely to be very low.

[sush]

Hello, I'm relatively new to bioinformatics analysis. I'm using MMSEQ for the first time, and I'm having trouble installing it properly (I'm using a Unix system).

I've downloaded and unzipped MMSEQ 0.11.2 from http://bgx.org.uk/software/mmseq_0.11.2a.zip. However, the command "make all" returns the following:

g++ -DVERSION=0.11.2 -fopenmp -O3 sokal.o uh.o hitsio.o mmseq.cpp -o mmseq -lgsl -lgslcblas -lboost_regex -lboost_iostreams -lbam -lz
/usr/bin/ld: cannot find -lbam
collect2: ld returned 1 exit status
make: *** [mmseq] Error 1

Any suggestions as to how to install properly?

[eturro]

The package comes with pre-built linux binaries which you should be able to run out of the box. No need to run make. The reason you're getting that error is that the linker can't find libbam.a (the samtools library), which mmseq depends on.