keep strand

[kullrich]

Hi, according to the manual at the moment, the reverse complement and in case of is_protein=True all 6 possible open reading frames are considered. Is there an option to keep the strand and not do reverse complement or just keep 3/6 of possible ORFs? Thank you in anticipation Best regards Kristian

[ctb]

hi @kullrich this is not possible at the moment but would not be hard for us to implement.

we could easily give you a script to produced sketches like this, and the rest of sourmash would work fine with the resulting sketch. interested? (It just will take time to add that feature into sourmash sketch.)

see also https://github.com/sourmash-bio/sourmash/issues/657

[kullrich]

Yes, I am interested. Actually, I am re-implementing this (https://github.com/mrvollger/StainedGlass) but using sourmash jaccard distances instead of blast . By this, the tool is much faster and can be used without a snakemake environment. Anyhow, I need to keep track of the strand to be able to color possible rearrangements accordingly. Best regards Kristian

[ctb]

neat! will see what I can do :)

[kullrich]

Here, e.g. chr4 as a comparison, but with jaccard distances. I just need to work on the color map and to implement strand specific scores. Even multithreading might be possible to calcualte pairwise jaccard distances in parallel (maybe with dask?)