Any Possible Method/Tool That Could Refers The Number Of Overlapping Sequences Used To Build Each Contig?

Question

[bambus0725]

HI,

I work with Metatranscriptomics data(sequenced using Illumina technology). I did de-novo transcriptome assembly using SOAP-Denovo-Trans assembler and now looking for a tool/software that could help me out to find the total number of overlapped reads involved to form a single contig to understand how good or bad the coverage is.

Any suggestions could be helpful.

Thank you in advance.

Answer 1

[Alex Reynolds]

If you have your contigs and reads in sorted BED format (e.g., myContigs.bed and myReads.bed) and you know your overlap criteria, then you could use BEDOPS bedmap to answer that question:

$ bedmap --echo --count <overlap-options> myContigs.bed myReads.bed > myContigsWithCountOfOverlappingReads.bed

If you leave out <overlap-options> then the default overlap between read and contig is one base. Otherwise, you can specify number of bases of overlap between files with --bp-ovr or require a fraction of contig or read length with --fraction-ref and --fraction-map respectively. Other overlap options are also available. This is discussed more fully in the BEDOPS documentation.

[bambus0725]

Thank you for the comment Alex.

the problem is that the data I work with is in Fasta format,is there an option to convert fasta file to the format that could be acceptable by BEDOPS like BAM/SAM.Does it works?

[Alex Reynolds]

It will depend on your Fasta file and whether it already contains coordinate and chromosome information (in the header, for instance). If not, you'll need to align your sequences to a reference genome to turn into BAM, SAM or PSL, and convert from there into BED with a conversion script (such as those in BEDOPS).

[bambus0725]

No actually the header line has only the sequence ID,and for most of the organisms doesn't exist any reference genome yet so I guess I can't do, even this is the reason I did de-novo transcriptome assembly