SEQanswers

SEQanswers (http://seqanswers.com/forums/index.php)
-   General (http://seqanswers.com/forums/forumdisplay.php?f=16)
-   -   Plant DNA barcoding markers (http://seqanswers.com/forums/showthread.php?t=75668)

lvogel 04-25-2017 05:45 AM

Plant DNA barcoding markers
 
Hi, I have a general question about DNA barcoding of plants (I just need a brief answer, not a lot of detail). I've heard/read that a combination of matK and rbcL sequences are customarily used to provide the highest resolution of classification, but even with these two, it isn't necessarily to species level. So I was wondering: Can both matK and rbcL be BLAST-ed, or do we need another method? And should you do a certain gene first, to get to e.g., family level, and then narrow it down afterwards, or can you just BLAST everything all at once?:confused:

lvogel 05-18-2017 03:36 AM

I have found a (partial) answer to my own question, by reading a paper involving barcoding of tea samples, which I'm also interesting in doing. It seems that it was enough to confirm the presence of any plant species if even one of the primers yielded sequences with a high enough % identity to references. (Mabye over 99.5%).
Please feel free to correct me if I'm wrong, though.

ATϟGC 05-18-2017 12:02 PM

As you have mentioned, plant DNA barcoding to the species level is rather imperfect to put it mildly. Incomplete lineage sorting and hybridization have led to the sharing of chloroplast haplotypes in many taxa. Even if you have complete chloroplast genomes species-level identification with DNA is not possible in many groups due to rampant haplotype sharing.

If you do not have very many sequences to identify and/or if the taxonomic range is small, it is probably preferable to use a cladistic approach to construct a phylogeny with closely related sequences. You could use sequences from your own custom reference library (e.g. species in your study area) or use sequences from a BLAST search. If you have a large number of sequences from a broad range of taxa you might want to use a more high-throughput method like the RDP-naive Bayesian classifier (Wang et al 2007) trained with your own custom database or some other sequence clustering method.


All times are GMT -8. The time now is 11:36 PM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2018, vBulletin Solutions, Inc.