Transcriptome reconstruction from both short reads and long sequences

Question

[Prakki Rama]

Hi all, Could I please know if there is any tool to reconstruct transcriptome from reference genome using both short read and long sequences from pacific biosciences at one go simulataneously.

I need something like Scripture, but it is limited to short reads. I would prefer a tool which can deal with both short and long at the same time. Thanks in advance for your suggestions.

Answer 1

[mark.ziemann]

How long are your reads?

You can use pre-aligned data (bam files) for scripture.

http://www.broadinstitute.org/software/scripture/Walkthrough_example

Use a dedicated long-read aligner to generate the alignment and then use Scripture to do the reconstruction. STAR might be a good option for alignment depending on error rates.

[Prakki Rama]

Thank you. I want to use both Illumina as well as pacbio data. The example shows only Illumina reads mapped to genome. So, you say, as long as the data is in pre-aligned format (bam files) it should be ok? My long read data mean length is 6.3 kb.

[mark.ziemann]

Recent versions of STAR (i.e., 2.4.1c) are distributed with STARlong that is optimised for reads >200bp in length. There isn't anything in the manual about it, but here is a comment from the author about it. You may want to use standard STAR for Illumina reads and STARlong for the PacBio reads.

Answer 2

[5heikki]

A de novo approach using idba_tran:

idba_tran -r $1 -l $2 -o $3 --num_threads 16 --mink 20 --maxk 100 --step 5
-r pe reads in interleaved fasta
-l long reads in fasta
-o output dir

[Prakki Rama]

Have not tried this. Will have to check. Thank you.

Answer 3

[CraigM]

How about MIRA?

An EST assembler which can perform hybrid assemblies using platforms of different read lengths.

I do not have first hand experience of trying this tool yet but believe it can do what you are looking for.

http://sourceforge.net/p/mira-assembler/wiki/Home/

A list of software for PacBio assembly, including hybrid assemblies, can be found here https://github.com/PacificBiosciences/Bioinformatics-Training/wiki/Large-Genome-Assembly-with-PacBio-Long-Reads

[Prakki Rama]

MIRA appears denovo assembler. I want to use genome as reference and run it. Will have check the tool. Thanks for the suggestion.