search

loneknightpy / idba

124 stars 53 forks source link

puzzled with the results of the idba_ud #34

Open sjtuyf opened 6 years ago

sjtuyf commented 6 years ago

Hello Yu Peng, The idba_ud tool is published by you.It is a so useful tool that I want use in my project.However,I got some problems in the process. My command is "idba_ud -r 01_merge.fa --pre_correction --min_contig 500 --step 10 --seed_kmer 55 --num_threads 9 --mink 52 --maxk 92 -o 02_idba_assem_2".The results include "contig-52.fa,contig-62.fa,contig-72.fa,contig-82.fa,contig-92.fa,contig.fa,scaffold.fa",I do not know which one I should use to regard as the final contigs ,in order to predict the genes.In other words,in which condition,I should use the contig.fa,in which condition I should use the scaffold.fa. Because it make me very puzzled for a long time,I want to get some help from you. I wish you can give me suggestions. The best wishes to you. I am looking forward to you reply.Thanks a lot.

loneknightpy commented 6 years ago

@sjtuyf Please use contig.fa and scallold.fa. All other files contain intermediate results. contig.fa contains the contigs without gaps. scallfold.fa is produced by joining contigs using paired-ends information and it may contain gaps.

sjtuyf commented 6 years ago 
 Ok,I feel excited to get your reply.Thank you very much.For my project,it is about the marine metagenome assemble.And next,I will use the contig.fa to predict the genes.So I think the step of assembling is important.
 Some other questions are about the parameter setting. Different papers use different parameters.I list some parameters in different papers.So which is better is a question.For the --seed-kmer,is it better for me to use it ?Or,it depends on the setting kmer?
 For the mink and maxk.I also try two times.And the results are in the attachment.Please find them in the attachment.
 In your opinion,which parameter setting is more better?Or in the references,which one I should follow?
 I wish you can give me some suggestions.Sorry to trouble you so much.Thanks again.

Best, Yafei idba_result.txt