Hello,
It would be great if the current version 9 of the RDP training could be modified to be compatible with mothur…
Thanks,
Tamar
Arggh, thanks for catching this. After going a few years with no updates they’ve put two out in 6 months. We’ll get on this ASAP.
Pat
See:
http://www.mothur.org/wiki/RDP_reference_files
For the update. Thanks again for the heads up.
Pat
Hi, will there also be a mothur-compatible training set of the RDP release 10 with the next version of mothur?
Or does this not add much?
Kind regards,
FM
As far as I can tell trainset9 is still current:
Use our classifier to assign 16S rRNA or Fungal LSU sequences to the new phylogenetically consistent higher-order bacterial and fungal taxonomy. Hierarchical taxa assignment is based on RDP naïve Bayesian rRNA Classifier. This is the current RDP Classifier Version 2.5 trained on 16S rRNA training set 9 and Fungal LSU training set 1. See more details about 16S taxonomy.
Sorry for wasting your time. I wasn’t aware there was no update in the training set for the classifier, which is apparently only updated for the Infernal aligner…
The hand-curated RDP Infernal Alignment training data for Bacteria and Archaea to build Infernal CM models is available.
These files reflect the contents of the current RDP Release 10 data.
http://rdp.cme.msu.edu/misc/resources.jsp;jsessionid=DEBA38C7988220FF1D6A9CA0223A7781.staghound#aligns
However the training set is not adapted for the NAST aligner which is implemented in mothur. I was just wondering whether that would be possible at all, but at the moment I am also using the Silva seed reference alignment which is available from the mothur website.
Kind regards.
FM
No, we’re not likely to include the rdp reference alignment since we don’t think this approach would make sense. The infernal alignment doesn’t align the most variable regions of the 16S gene. You really would be better off just using the SILVA reference alignment that we posted.
Hi Pat,
It looks like RDP has just updated to version 2.6 and added a fungal training set.
“This is the current RDP Classifier Version 2.6 trained on 16S rRNA training set 9 and Fungal LSU training set 8.”
Will mothur be adding these?
Thank you,
Kathie Mihindukulasuriya
Hi Kathie,
The training set that we provide at http://www.mothur.org/wiki/RDP_reference_files is the 16S training set v9. And what we have as Fungal LSU training set 7 seems to be the same as Fungal LSU training set 8. Everything seems to be up to date.
Pat
Thank you very much Pat.
Dear Patrick,
I believe there has been an update of the RDP for 16S, could you please adapt it for Mothur use?
http://rdp.cme.msu.edu/misc/resources.jsp
http://rdp.cme.msu.edu/misc/rel10info.jsp
Version 11
Thanks a lot!, Best Regards,
Rudiger.
Actually, looking deeper it seems that the classifier training set is still version 9, which is what we have up.
Pat
Hi sorry for the most likely stupid question but the RDP training set vs 14 you provide does not seem to be in the format used in mothur. It gives the error below.
[ERROR]: 0Root-10rootrank is missing the final ‘;’, ignoring.
[ERROR]: 2*“Actinobacteria"12phylum is missing the final ‘;’, ignoring.
[ERROR]: 4Acidimicrobidae34subclass is missing the final ‘;’, ignoring.
[ERROR]: 6"Acidimicrobineae"56suborder is missing the final ‘;’, ignoring.
[ERROR]: 8Acidimicrobium78genus is missing the final ‘;’, ignoring.
[ERROR]: 10Ferrithrix78genus is missing the final ‘;’, ignoring.
[ERROR]: 12Iamiaceae67family is missing the final ‘;’, ignoring.
[ERROR]: 14"Acidimicrobineae”_incertae_sedis67*family is missing the final ‘;’, ig and
DQ343153|S000640727’ is in your template file and is not in your taxonomy file. Please correct.
‘EU928765|S001872839’ is in your template file and is not in your taxonomy file. Please correct.
‘AY639887|S000333610’ is in your template file and is not in your taxonomy file. Please correct.
‘EU167539|S001044475’ is in your template file and is not in your taxonomy file.
I see that you have a readme on how you converted the reference files. I just assumed the copy online was already converted or must I follow the readme page?
Sorry stupid of me I never extracted the compressed file on my terminal. It is working fine now. :oops:
Hi there,
Some of my colleagues have been using the RDP version 7 fungal 28S files that were made available on the wiki. The version 7 release contains 8506 sequences. http://mothur.org/wiki/RDP_reference_files
I happened to look at the RDP page the other day and noticed that the May 2015, Release 11.4 includes 108, 901 fungal 28S rRNA sequences. Do you anticipate files for fungal 28S sequences from this release becoming available on the wiki?
Keep in mind that there’s a difference between the training sets and the database sequences. The training sets are curated while the broader database is not as well curated. Also, fungi aren’t really our game - so if you or anyone else would like to generate a database and post it to the wiki, that would be awesome.
Pat