I would like to ask you for help with interpretation of abgd results. According to morphology and phylogenetical tree clades, I should have here 6 species. ABGD gives me some weird results, with no barcoding gap, moreover I don't know how to connect the plot with partitions with the histogram of distances, because the scales are in different orders of magnitude. Anyone has some ideas what's wrong here?
The method is nice, but it is not a dogma. For example, several species are not recognized by barcoding, yet they are valid. This is especially frequent in freshwater fishes; research the case of the Cyprinodon species flock of Lake Chichancanab, Mexico, where 4 out of 7 species have no exclusive haplotype, yet their striking morphological differences are genetically based and ecologically relevant.
Hi Beniamin, have you tried to run your analysis at this link for the COI?
http://wwwabi.snv.jussieu.fr/public/abgd/abgdweb.html
It worked very well with my analysis.
If you have tried already, can you see a barcoding gap between your species when you compare the COI genetic distance? (e.g. after alignment, if you create a pairwise distance matrix?)
Cheers
Francesco
Thank you for answers!
Yes, those results are from this website. I did not try to detect barcoding gap in distance matrix, but according to phylogenetical tree, there are well-supported clades that correspond with the morphological species. It does not work for the COI, because there was probably introgression, but for both nuclear genes, results are congruent with the morphology.
And I understand that this method is not a dogma, however I would like to somehow explain these strange results.
Best regards,
Beniamin
That's quite interesting.
Usually, barcoding gap is calculated using COI. Issues with your COI sequences may lead to issues with ABGD. I would suggest you to try and calculate the barcoding gap using just the COI and a distance matrix in order to compare the results.
Moreover, you may want to check if your sequences are saturated (which may again lead to these being useless for phylogenetic studies). A nice software to check saturation is DAMBE (references: XIA 2008 and XIA 2013).
Keep us posted with your results, and good luck!
This method sort sequences into putative species based on pairwise distance, you can work with distance matrix or you can use. the alignment that the ABDG web serve make the distance matrix before analyzes. The ABDG web serve separates the species based on a range of maximum intraspecific distance. I mean, the ABDG consider different value of maximum intraspecific divergence to separete species (graph below) .You can click each point (yellow or red) to see which species were delimited for each maximum intraspecific divergence. If you know the intraspecific divergence of the gene in your taxonomic group, you can modify the parameter on web serve and maybe you get better results. Other important point is: the ABDG works better if you use the sequences closer phylogenetic each others(for e.g use sequences from the same class or genus). Because the substitutions rates (consequently intraspecific distance ) can vary among taxonomy groups. For this reason, I recommend exclude the outgroup from alignment before analyses. Since the greater pairwise distance between ingroup and outgroup can "confuse " ABDG.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs