very weird alignment and results

[marlenec]

Hi,
I am new using this software. I have an Illumina sequencing of V4 sequence 16S rDNA. I trimmed and assembled the sequences using DADA2.
I wanted to use hmmufotu on DADA2's ESVs to compare the taxonomic classification of the two softwares, and get a phylogenetic tree for my ESVs.

Here is the beginning of the input :

Otu1
AGCAGTGGGGAATAT[...]CAAACAGGATTAGATACCCTGGTA
Otu2
AGCAGTGGGGAATAT[...]GGATTAGATACCCTGGTA
Otu3
AGCAGTGGGGAATAT[...]AGGATTAGATACCCTGGTA

After running hmmufotu and hmmufotu-sum on the file using GreenGenes (v13.8) species-level (97% OTU) reference + GTR DNA model that is recommanded, I got a very weird alignment were almost all bases of most of my ESVs (here they are called Otus, but it is only for compatibility with other software) are replaced by gaps '-'

Example

5360 DBName=Archive/GTR/gg_97_otus_GTR;Taxonomy="k__Bacteria;p__Firmicutes;c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__[Ruminococcus];s__gnavus";AnnoDist=0.64307999999999976;ReadCount=71;SampleHits=1
--------------[...]----------------------------------------------
--------------------------TACCAGGGCTACACACGTGCT----
---[...]-----

[...] are were I reduced the sequence length for the purpose of this message)

And the classification is also very weird, with only 7% of agreement at Phylum level with classification on SILVA database using RDP classifier.

Perhaps I am using it wrong ? I know this software is supposed to be used on raw reads, but I thought it would have been great to compare its classification resolution with RDP classifier.

Thanks you in advance!

[e00011027]

Hi, Thanks for using HmmUFOtu. Most of this type of problem is you are using an older versio of HmmUFOtu, and failed to tell the correct strand-ness of the reads. In other words, your fwd/rev reads were in wrong orientation (due to primer mis-naming). Try to upgrade to the latest HmmUFOtu, and either let it "guess" the correct strand (by default), or manually tell it the strand orientation. Check -h help message for details. Best Qi

…

On Thu, Aug 8, 2019 at 12:41 PM marlenec ***@***.***> wrote: Hi, I am new using this software. I have an Illumina sequencing of V4 sequence 16S rDNA. I trimmed and assembled the sequences using DADA2. I wanted to use hmmufotu on DADA2's ESVs to compare the taxonomic classification of the two softwares, and get a phylogenetic tree for my ESVs. Here is the beginning of the input : Otu1 AGCAGTGGGGAATAT[...]CAAACAGGATTAGATACCCTGGTA Otu2 AGCAGTGGGGAATAT[...]GGATTAGATACCCTGGTA Otu3 AGCAGTGGGGAATAT[...]AGGATTAGATACCCTGGTA After running hmmufotu and hmmufotu-sum on the file using GreenGenes (v13.8) species-level (97% OTU) reference + GTR DNA model that is recommanded, I got a very weird alignment were almost all bases of most of my ESVs (here they are called Otus, but it is only for compatibility with other software) are replaced by gaps '-' Example 5360 DBName=Archive/GTR/gg_97_otus_GTR;Taxonomy="k__Bacteria;p__Firmicutes;c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__[Ruminococcus];s__gnavus";AnnoDist=0.64307999999999976;ReadCount=71;SampleHits=1 --------------[...]---------------------------------------------- --------------------------TACCAGGGCTACACACGTGCT---- ---[...]----- [...] are were I reduced the sequence length for the purpose of this message) And the classification is also very weird, with only 7% of agreement at Phylum level with classification on SILVA database using RDP classifier. Perhaps I am using it wrong ? I know this software is supposed to be used on raw reads, but I thought it would have been great to compare its classification resolution with RDP classifier. Thanks you in advance! — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#5?email_source=notifications&email_token=ABXZ44L7GEVEDY7PKMO4T3LQDREEDA5CNFSM4IKMPD5KYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HEGRK3Q>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABXZ44JZ7DE6WSQCVDG5HE3QDREEDANCNFSM4IKMPD5A> .

[marlenec]

Hi, Thank you for your answer. I tried with -t option and it worked. Thanks, Marlène

…

----------------------------------------------- Dr Marlène Chiarello Dpt of Biology University of Mississippi Oxford MS, 38655 USA Cell (France): +33 (0)6-72-13-72-12 Cell (USA): +1 (662) 202-7433

Le 8 août 2019 à 13:12, Qi ***@***.***> a écrit : Hi, Thanks for using HmmUFOtu. Most of this type of problem is you are using an older versio of HmmUFOtu, and failed to tell the correct strand-ness of the reads. In other words, your fwd/rev reads were in wrong orientation (due to primer mis-naming). Try to upgrade to the latest HmmUFOtu, and either let it "guess" the correct strand (by default), or manually tell it the strand orientation. Check -h help message for details. Best Qi On Thu, Aug 8, 2019 at 12:41 PM marlenec ***@***.***> wrote: > Hi, > I am new using this software. I have an Illumina sequencing of V4 sequence > 16S rDNA. I trimmed and assembled the sequences using DADA2. > I wanted to use hmmufotu on DADA2's ESVs to compare the taxonomic > classification of the two softwares, and get a phylogenetic tree for my > ESVs. > > Here is the beginning of the input : > > Otu1 > AGCAGTGGGGAATAT[...]CAAACAGGATTAGATACCCTGGTA > Otu2 > AGCAGTGGGGAATAT[...]GGATTAGATACCCTGGTA > Otu3 > AGCAGTGGGGAATAT[...]AGGATTAGATACCCTGGTA > > After running hmmufotu and hmmufotu-sum on the file using GreenGenes > (v13.8) species-level (97% OTU) reference + GTR DNA model that is > recommanded, I got a very weird alignment were almost all bases of most of > my ESVs (here they are called Otus, but it is only for compatibility with > other software) are replaced by gaps '-' > > Example > > 5360 > DBName=Archive/GTR/gg_97_otus_GTR;Taxonomy="k__Bacteria;p__Firmicutes;c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__[Ruminococcus];s__gnavus";AnnoDist=0.64307999999999976;ReadCount=71;SampleHits=1 > --------------[...]---------------------------------------------- > --------------------------TACCAGGGCTACACACGTGCT---- > ---[...]----- > > [...] are were I reduced the sequence length for the purpose of this > message) > > And the classification is also very weird, with only 7% of agreement at > Phylum level with classification on SILVA database using RDP classifier. > > Perhaps I am using it wrong ? I know this software is supposed to be used > on raw reads, but I thought it would have been great to compare its > classification resolution with RDP classifier. > > Thanks you in advance! > > — > You are receiving this because you are subscribed to this thread. > Reply to this email directly, view it on GitHub > <#5?email_source=notifications&email_token=ABXZ44L7GEVEDY7PKMO4T3LQDREEDA5CNFSM4IKMPD5KYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HEGRK3Q>, > or mute the thread > <https://github.com/notifications/unsubscribe-auth/ABXZ44JZ7DE6WSQCVDG5HE3QDREEDANCNFSM4IKMPD5A> > . > — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#5?email_source=notifications&email_token=AFUPUCQQLYLYWSSJTWYYUXTQDRO2TA5CNFSM4IKMPD5KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD34OX2Y#issuecomment-519629803>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFUPUCWURHTKW3XXSN5JN7LQDRO2TANCNFSM4IKMPD5A>.