Solid Split Rna-Seq Mapping

Question

[Darked89]

I got few SOLID RNA-Seq files which I am trying to map to some "draft+ quality" mammalian genome. So far I tried:

• bowtie 0.12.7 (unspliced mode)
• tophat 1.3.1 (spliced)
• abwtp1.2.1 (spliced)

Bowtie works OK but obviously misses spliced reads. Tophat without annotations maps less reads than bowtie. Whole Transcriptome Pipeline gives me colorspace gff file which then GffToSam fails to convert to SAM (see below). In general tools from http://solidsoftwaretools.com/ have this abandon-ware feel about them, with discussion forums with year+ old unanswered messages, README files pointing to i.e. g++ 3.4.5, and compilation errors due to g++ changes.

Hence my question: what do you use for spliced color space RNA-Seq mapping?

re: GffToSam

./GffToSam -i ./ab_wtp.110811.OUT/output/wta_1_rnaA_f3.sorted.max.merged.filtered.gff  -o result.sam -id S1 -sm TEST

gives:

eq. name at line "MT    wtp read    3326    3368    39  +   .   bd=1657_1171_1518_F3;rs=1;mm=2;g=T32202210023310223200333030123302000002103000103202;i=1;" doesn't have F3 or R3 tag. Skipping.

[Istvan Albert]

You have two different questions here and that makes answering them more difficult.

The GFF to SAM conversion fails because that is not a valid operation to being with. A GFF file is not expected to contain the information needed to produce a SAM file. The only case this would work if someone artificially inserted certain attributes into the GFF file to make it more like a SAM (therefore creating a frankensteinian GFF that is really a SAM file).

[Darked89]

@Istvan: as far as I know GFF is the default output of few other SOLID mapping tools. So using them implies that GffToSam works for some people, or they are not using any GFF-producing SOLID mapping software (for various reasons). I bumped into two roadblocks (tophat, ab-wtp), so I am interested in working solutions. Hence 'combined' question.