search

mrmckain / FASTKs

Pipeline for generate Ks plots of transcriptome data.
13 stars 6 forks source link

problems with interpretation of data #3

Open kiraza-lab opened 3 years ago

kiraza-lab commented 3 years ago

Dear Michael, thank you for the tool! In present I'm trying to compare the data for species from the same genus. Based on my previous results, one of them (species#4) underwent a recent WGD, while othes (species#1-3) are considered to be pre-WGD species. I'm not sure how to interpret the plots I obtained with FASTKs. I mean the height of the purple and black peak. When it means pure WGD event or just small tandem duplications? Can I say anything regarding the age of the duplications (young or old)? Thank you so much for the help! Kira. plots forum

mrmckain commented 3 years ago

One issue with FastKs is that it will sometimes have so many pairs that events might get obscured. Makenzie Mabry showed a comparison of FastKs and the Barker pipeline in her recent paper (https://bsapubs.onlinelibrary.wiley.com/doi/full/10.1002/ajb2.1514) where this happens. The colored lines are from mclust, which will over group subsets and need some fiddling with the parameters at times.

Looking at these graphs, it is hard to say there is a WGD in any of them. The 4th one has a lot of very similar putative paralogs, which might suggest there was a recent duplication, but when the events are young and the subgenomes not very diverged, it is very hard to say for sure. As for age, you can't compare for age across different Ks plots. Rate variation can obscure things. I did some things things in this paper (https://pubmed.ncbi.nlm.nih.gov/22301890/) to alleviate this to identify when an event occurred relative to speciation events. I usually try rely on multiple methods (Ks, phylogenies, synteny if I have genomes) to gather evidence when drawing conclusions of when an event might have occurred.

Best, Michael

kiraza-lab commented 3 years ago

Dear Michael, thank you for the detailed answer. In the present, I don't have genomes of these species (except for the last one), and now it's clear that I need to use additional methods for correct interpretation of the Ks plots. Thank you so much for the help! Best wishes, Kira.