dr-artio / KGEM

KGEM: Scala implementation of Reconstruction of viral population algorithm for single-amplicon sequencing data
http://alan.cs.gsu.edu/NGS/?q=content/kgem
2 stars 2 forks source link

Creates empty file #3

Open averagehat opened 9 years ago

averagehat commented 9 years ago
bash-4.1$ java -cp target/kgem-0.7.1.jar:target/dependency/* edu.gsu.cs.kgem.exec.Main   780.sam  1 && head haplotypes.fa 
[main] INFO kGEM - [04:49:55 PM] -----------------------------------------------------------
[main] INFO kGEM - [04:49:55 PM] kGEM v.null: Local Reconstruction for Mixed Viral Populations.
[main] INFO kGEM - [04:49:55 PM] -----------------------------------------------------------
[main] INFO kGEM - [04:49:55 PM] Reading input...
(ngs_mapper) bash-4.1$ 

Here is the file: https://github.com/averagehat/biopandas/blob/9f263702f967ccbefd888b837f9122767806106c/780.sam

I also tried a file with only one reference sequence but that also failed. Additionally, if I try to run with optional arguments, aka

 java -cp target/kgem-0.7.1.jar:target/dependency/* edu.gsu.cs.kgem.exec.Main   out.sam  -k 50 -tr 5

I just get the usage echoed back to me.

Thanks

dr-artio commented 9 years ago

Currently kGEM accepts as input only aligned reads in fasta format. This format is spanning and extending reads with '-' instead of SAM records. For conversion required reference and sam file and https://github.com/night-stalker/ERIF tool It is not very handy but parameters -sam reads.sam -g ref.fasta -o output.fasta I am working on merging those codes so kGEM will be able to accept SAM file as input but this will be added only in future release.

averagehat commented 9 years ago

I see, I got confused by .sam extension in the example of usage:

java -jar KGEM.jar reads.sam -k 50 -tr 5

I will try out the ERIF tool.

dr-artio commented 9 years ago

Yea I see, README haven't been updated for a while, anyway it will support that later, also note that -k is not an argument so usage is kgem reads.sam 50 -o out.fas this k is not optional now. Also use help usage as a guide since README is outdated a bit.