Report all mappings

[Donovan Parks]

Hi Philip,

I’m having trouble getting KMA to report all mappings. I have a simple test case (see attached) that contains two identical reference sequences and 3 query sequences that map to these reference sequences at 100% identity. From the KMA output in *.frag.gz I can see that it identifies 2 best mappings for each of the query sequences, but only one mappings is reported in the *.frag.gz file. The same is true if I output a SAM file. Only 1 of the two best mappings are reported.

I am running KMA as follows:

kma -a -sam -i query.fna -o mappings -t_db ref

My understanding is that the -a flag should cause all template mappings to be reported. Are there other parameters I need to set to make this happen?

Thanks,

Donovan

[ptlcc]

Hi Donovan

The multi mapping information will be present in the mappings.frag_raw.gz file, which follows the format of mappings.frag.gz, with lists over single values however.

There was a bug in the reporting of this in version 1.3.25, this has been fixed in 1.3.26.

Best,
Philip

[alvanuffelen]

Hi Philip

I am having trouble identifying the best mappings of a given query.
I enabled the -a option, so a *.frag_raw.gz file was generated. I can find all the multi-mappings for a given query but I do have a few questions:

1. In the *.frag_raw.gz, the template names have been number encoded according to *.name index file. However, *.frag_raw.gz reports many zeroes as index for a template. What is the meaning of a zero in the template column (see ‘template_frag_raw.gz’ column)?
2. How do I identify the ‘equally well mapped’ templates from *.frag.gz? *.frag.gz reports only one explicitly and the number of ‘equally well mapped’ templates is not the same as the number of multi-mappings of *.frag_raw.gz (see image). Gave a closer look to your paper. *.frag.gz reports number of best-aligning templates using ConClave score while *.frag_raw.gz just reports the number of best-aligning templates based on alignment score.

I attached an image for clarification. The image shows a subset of an union of *.frag_raw.gz and *.frag.gz on the ‘query’ column. All columns left of ‘query’ are from *.frag.gz and right of ‘query’ are from *.frag_raw.gz

‌

[ptlcc]

Hi Al

It seems that the column has been swapped with a position column. The frag files might contain two extra fields if the query was hard clipped, which might have caused the shift in the table.

Best,
Philip