Loading...
 

Support Help

Forums > Support> Seemingly unique hits not reported...

Seemingly unique hits not reported...

Hi,

I mapped several simulated reads with novoalign in its default reporting mode (-o SAM -o fullNW -e 1 --ILQ_QC) and asking it to report multiple hits (-o SAM -o fullNW -r A 20 -q 5). I got several reads for which the multiple hit mode gives only one hit, and the default mode does not report it. indicating multiple hits. I have attached an example of such reads. Can you please explain the reason for this? Could it be that novoalign does not report multiple hits if there are too many of them?

Mapping reported under default mapping
893 4 * 0 0 * * 0 0 caataaacatatgtgtgcatgtgtctttatagcagcatgatttatagtcctttgggtatatacccagagagggga >=:7@>7=>??4'>8D9?6'3@C5;4?>68.@5=:7@>7=>??4'>8D9?6'3@C5;4?>68.@5


Sorry, noticed that the reads were not posted correctly. I have attached them in a text file..
> Hi,
>
> I mapped several simulated reads with novoalign in its default reporting mode (-o SAM -o fullNW -e 1 --ILQ_QC) and asking it to report multiple hits (-o SAM -o fullNW -r A 20 -q 5). I got several reads for which the multiple hit mode gives only one hit, and the default mode does not report it. indicating multiple hits. I have attached an example of such reads. Can you please explain the reason for this? Could it be that novoalign does not report multiple hits if there are too many of them?
>
> Mapping reported under default mapping
> 893 4 * 0 0 * * 0 0 caataaacatatgtgtgcatgtgtctttatagcagcatgatttatagtcctttgggtatatacccagagagggga >=:7@>7=>??4'>8D9?6'3@C5;4?>68.@5=:7@>7=>??4'>8D9?6'3@C5;4?>68.@5


Hi drcyber,

Could you post full command line you are using, information regard reference genome and the version of Novoalign you are using. If not latest version can you try with latest version.

Thanks, Colin


Hi Colin,

I am using the latest Novoalign V2.08.01

the genome is hg19.

I used the command
novoalign -d hg19_novoIndex -f simulated_reads.fq -o SAM -o fullNW -r A 20 -q 5 --ILQ_QC
for multiple reads,
novoalign -d hg19_novoIndex -f simulated_reads.fq -o SAM -o fullNW -e 1 --ILQ_QC
for default mapping..

Thanks


Hi,

The problem is the -e 1 option, take that off and all should be OK.

The -e option is design to reduce time taken to align reads that hit high copy number repeats. It stops the alignment process as soon as we have found n alignments equal to the best one already found. If you set it too low it could stop the alignment process before the best alignment was found. I'd never use it adds a heuristic and increases the chance of false positive alignments. On occasions I have used I set it fairly high to ensure its an alignment to a high copy number repeat.

It can improve performance a bit bt it does mean reads that align to repeats might not get aligned in the best location. Setting -e 1000 or so should be safe.

Kind Regards, Colin


Yes, It solves the issue, Thanks a lot.

Show posts:
 
Show HelpHelp