[khmer] Fwd: About Khmer and Trinity.

Won C Yim wyim at unr.edu
Thu Jan 16 11:01:49 PST 2014


Dear Brown,

Hi!

I have question.

I have a 114 reads data.

First I ran Trinity with whole dataset with adapter and quality trimmed fastq.

The no. of contigs were 108,112 .

After your normalization it generated 22,454 contigs.

I know it works well but our species has 295Mb.

The Ararbidopsis is 135Mb and have 30,000 genes.

I think 22,454 contigs were to small.

Do I lost something?

Please help.

Thank you.


> 보낸 사람: "Yim,Won Cheol" <wyim at unr.edu>
> 제목: Re: About Khmer and Trinity.
> 날짜: January 15, 2014 at 2:02:35 PM PST
> 받는 사람: "<titus at idyll.org>" <titus at idyll.org>
> 
> Dear Brown,
> 
> Hi!
> 
> I have question.
> 
> I have a 114 reads data.
> 
> First I ran Trinity with whole dataset with adapter and quality trimmed fastq.
> 
> The no. of contigs were 108,112 .
> 
> After your normalization it generated 22,454 contigs.
> 
> I know it works well but our species has 295Mb.
> 
> The Ararbidopsis is 135Mb and have 30,000 genes.
> 
> I think 22,454 contigs were to small.
> 
> Do I lost something?
> 
> Please help.
> 
> Thank you.
> 
> Won
> __
> Yim, Won Cheol
> 
> Postdoctoral Fellow
> 
> MS330/Department of Biochemistry & Molecular Biology
> 1664 N. Virginia Street
> University of Nevada
> Reno, NV 89557-0330
> 
> Mobile:775-229-0453
> Email: wcyim at unr.edu
> http://ascendo.u.googlepages.com
> 
> Jan 10, 2014, 5:29 AM, C. Titus Brown <ctb at msu.edu> 작성:
> 
>> Hi Won, the k-mer size is not directly analogous.  We've had fine experience
>> with k=20.
>> 
>> best,
>> --titus
>> 
>> On Fri, Jan 10, 2014 at 06:19:21AM +0000, Won C Yim wrote:
>>> Dear Titus,
>>> 
>>> Hi!
>>> 
>>> I have a question for your tutorial.
>>> 
>>> I think Trinity use 25mer for assembly.
>>> 
>>> Is it Okay to use 20mer for kmer?
>>> 
>>> Thank you.
>>> 
>>> Won
>>> __
>>> Yim, Won Cheol
>>> 
>>> Postdoctoral Fellow
>>> 
>>> MS330/Department of Biochemistry & Molecular Biology
>>> 1664 N. Virginia Street
>>> University of Nevada
>>> Reno, NV 89557-0330
>>> 
>>> Mobile:775-229-0453
>>> Email: wcyim at unr.edu<mailto:wcyim at unr.edu>
>>> http://ascendo.u.googlepages.com
>>> 
>>> 2013. 12. 18., ???? 10:45, C. Titus Brown <ctb at msu.edu<mailto:ctb at msu.edu>> ????:
>>> 
>>> Hi Won,
>>> 
>>> https://khmer-protocols.readthedocs.org/en/v0.8.4
>>> 
>>> best,
>>> --titus
>>> 
>>> On Wed, Dec 18, 2013 at 06:38:18PM +0000, Won C Yim wrote:
>>> Dear Titus Brown,
>>> 
>>> Hi !
>>> 
>>> I am Won Cheol Yim in University of Nevada, Reno.
>>> 
>>> Recently, I try to de novo assemble one of plant species (~350Mb) by using RNA-seq data.
>>> 
>>> I already assembled by Trinity alone and Trinity normalization program.
>>> 
>>> But the result is too bad.
>>> 
>>> Actually we have deep sequencing (250GB of 100bp Illumina single read).
>>> 
>>> So there is so many weird contigs.
>>> 
>>> I read many things about that like cut-off based on FPKM and Isoform percent.
>>> 
>>> However there??s no answer.
>>> 
>>> I think Khmer can help my dataset.
>>> 
>>> Is it possible to use Khmer with Trinity.
>>> 
>>> There is no reference paper.
>>> 
>>> And Trinity use 25 kmer as default.
>>> 
>>> Can you advice me the critical options?
>>> 
>>> I??ve used ?? normalize-by-median.py -k 20 -C 20 --hashsize 2e9 --n_hashes 4 all.fastq  ??
>>> 
>>> My fastq file is already trimmed by Q>20 and illumina adaptors.
>>> 
>>> Would you mind advice me some helpful comment?
>>> 
>>> Thank you.
>>> 
>>> Won
>>> 
>>> 
>>> 
>>> __
>>> Yim, Won Cheol
>>> 
>>> Postdoctoral Fellow
>>> 
>>> MS330/Department of Biochemistry & Molecular Biology
>>> 1664 N. Virginia Street
>>> University of Nevada
>>> Reno, NV 89557-0330
>>> 
>>> Mobile:775-229-0453
>>> Email: wcyim at unr.edu
>>> http://ascendo.u.googlepages.com
>>> 
>>> 
>>> --
>>> C. Titus Brown, ctb at msu.edu
>>> 
>> 
>> -- 
>> C. Titus Brown, ctb at msu.edu
> 



More information about the khmer mailing list