See the debate on how to infer RNA virus origins from phylogenies: https://mbio.asm.org/content/10/2/e00289-19 … and the response. Interested parties should definitely check out the original alignments (especially Data Set S2) herehttps://mbio.asm.org/content/9/6/e02329-18 …
-
More seriously, I am concerned that there is an increasing trend to infer phylogenies on highly divergent viruses that share almost no sequence similarity and where the alignment is dubious at best. Beware of crap alignments.
Totally agree - problem is that no matter the MSA you *always* get an answer - the art is in determining whether you can trust that answer or not. Which, unfortunately, isn't easy. We need a new tool - AlCrap - it'll tell you whether your alignment is crap or not...
-
-
When the alignment is challenging, it should be estimated jointly with the tree (with appropriate statistical procedures to test fit, measure uncertainty). There are methods for this
#bali-phy. -
Beyond challenging in this case. Take a look.
- Nog 2 antwoorden
Nieuw gesprek -
-
-
With very divergent (or non-homologous) sequences the guide tree that the MSA software creates based on pairwise alignments that matters will fix the tree. The MSA is then built to support that tree.
-
Here is an experiment: generated 7 random amino acid sequences (1K long). The first bit looks like this:pic.twitter.com/TL86QSfp0L
- Nog 4 antwoorden
Nieuw gesprek -
-
-
Interesting because we have struggled with poor sequence alignments for picobirnaviruses: we showed this in our own modest (& hopefully not too incorrect!): but an issue is we don’t know what the solution is even within this single familyhttps://www.mdpi.com/1999-4915/10/12/685 …
-
Exactly! It can be hard enough within single families, but across all families?
Einde van gesprek
Nieuw gesprek -
Het laden lijkt wat langer te duren.
Twitter is mogelijk overbelast of ondervindt een tijdelijke onderbreking. Probeer het opnieuw of bekijk de Twitter-status voor meer informatie.