Hello,

I have been trying to use makeTranscriptDbFromGFF() to read my gff file but I get the following error:

extracting transcript information
Error in .prepareGFF3TXS(data, useGenesAsTranscripts) :
  Unexpected transcript duplicates
In addition: Warning message:
In .local(con, format, text, ...) :
  gff-version directive indicates version is   3, not 3

I have read that this was a problem when he file does not have transcript features. My file does and I have checked all the transcript IDs are unique. This error did not happen in Bioconductor 2.14 but appeared when I upgraded to 3.0. useGenesAsTranscripts=TRUE does not help.

The file I am using seems legitimate, its has been validated by genometools gt gff3, which is pretty strict. I am including my R code, session info and the two genes from my gff3 (if I use just them everything is fine, but using the full file causes a problem, although all genes are formatted in the same manner, I am afraid in order to reproduce the error I would have to share the entire gff3 file).

txdb <- makeTranscriptDbFromGFF("annotation.gff3")

> sessionInfo()
R version 3.1.0 RC (2014-04-05 r65382)
Platform: x86_64-pc-linux-gnu (64-bit)

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C
 [9] LC_ADDRESS=C               LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] parallel  stats     graphics  grDevices utils     datasets  methods
[8] base

other attached packages:
 [1] VariantAnnotation_1.10.5 Rsamtools_1.16.1         Biostrings_2.32.1
 [4] XVector_0.4.0            GenomicFeatures_1.16.3   AnnotationDbi_1.26.1
 [7] Biobase_2.22.0           GenomicRanges_1.16.4     GenomeInfoDb_1.0.2
[10] IRanges_1.22.10          BiocGenerics_0.10.0

loaded via a namespace (and not attached):
 [1] biomaRt_2.18.0     bitops_1.0-6       BSgenome_1.30.0    DBI_0.3.1
 [5] RCurl_1.95-4.3     RSQLite_0.11.4     rtracklayer_1.22.7 stats4_3.1.0
 [9] tools_3.1.0        XML_3.98-1.1       zlibbioc_1.8.0

head -n 100 annotation.gff3

##gff-version   3

C1      JCVI    gene    1449    1625    .       -       .       ID=Bo1g001000;Note=cyclic nucleotide gated channel
C1      JCVI    mRNA    1449    1625    .       -       .       ID=Bo1g001000.1;Parent=Bo1g001000;Note=cyclic nucleotide gated channel
C1      JCVI    exon    1449    1625    .       -       .       ID=Bo1g001000.1_exon_1;Parent=Bo1g001000.1
C1      JCVI    CDS     1449    1625    .       -       0       ID=Bo1g001000.1_cds_1;Parent=Bo1g001000.1
###
C1      JCVI    gene    1711    2833    .       +       .       ID=Bo1g001010;Note=Amino acid transporter family protein
C1      JCVI    mRNA    1711    2833    .       +       .       ID=Bo1g001010.1;Parent=Bo1g001010;Note=Amino acid transporter family protein
C1      JCVI    exon    1711    1799    .       +       .       ID=Bo1g001010.1_exon_1;Parent=Bo1g001010.1
C1      JCVI    CDS     1711    1799    .       +       0       ID=Bo1g001010.1_cds_1;Parent=Bo1g001010.1
C1      JCVI    exon    1937    2288    .       +       .       ID=Bo1g001010.1_exon_2;Parent=Bo1g001010.1
C1      JCVI    CDS     1937    2288    .       +       1       ID=Bo1g001010.1_cds_2;Parent=Bo1g001010.1
C1      JCVI    exon    2366    2833    .       +       .       ID=Bo1g001010.1_exon_3;Parent=Bo1g001010.1
C1      JCVI    CDS     2366    2833    .       +       0       ID=Bo1g001010.1_cds_3;Parent=Bo1g001010.1

Hello,

It looks like you are using old versions of R/GenomicFeatures. My first thought would be to update R to 3.1.2, and then update bioconductor:

source("http://bioconductor.org/biocLite.R")
biocLite()

Also, are the gene IDs unique as well? E.g.

Do you get the same results for:

    grep '      gene    ' annotation.gff | grep -o "ID=[^;]*;" | sort | wc -l

and

    grep '      gene    ' annotation.gff | grep -o "ID=[^;]*;" | sort | uniq| wc -l

(Use cntrl-v + 'tab' to insert tabs before and after "gene" in the regex.)

Hi Agnieszka,

In BioC 3.1 (the current devel version, will be released in April), the GenomicFeatures package has a new function: makeTxDbFromGRanges(). This is still a work-in-progress (the man page will be added soon). Can you please try it? You'll need to install R-3.2 (current devel version of R) first, then:

source("http://bioconductor.org/biocLite.R")
biocLite(GenomicFeatures)

Then import your file first as a GRanges object with:

library(rtracklayer)
gr <- import("path/to/your/file")

And finally do:

makeTxDbFromGRanges(gr)

If that still doesn't work, or if you cannot get R-3.2, then it would be great if you could give us access to your file so we can troubleshoot.

The plan is to switch the current implementation of makeTxDbFromGFF() to make it use makeTxDbFromGRanges() internally in the near future.

Thanks,

H.

Hi Agnieszka,

I had a similar problem recently and my solution was much more mundane. makeTranscriptDbFromGFF() is apparently very strict in the gff-version pragma. Thus, it want only a single space between the pragma and the number as in "##gff-version 3". However, the  sequence ontology web site where the format is defined is mute on the number of spaces between  the pragma and the number and my GeneDB gff files have three spaces.

In summary, just edit the gff file and remove some spaces did the trick for me. Hope this helps for you too,

Tjeerd

Hello,

Thank you very much for your responses. In my case it seems that upgrade to R/3.1.2 and fresh installation of bioconductor solved the issue. I have upgraded on my personal windows 8 machine, I am still waiting for the upgrade on the server I was using (don't have root permissions). I will report back if I encounter further problems. So far I have this:

> txdb <- makeTranscriptDbFromGFF("C:\\Users\\Aga\\Documents\\bioinf\\brassica\\thesis\\snpeff\\variantannotation.gff3\\variantannotation.gff3")
extracting transcript information
Extracting gene IDs
extracting transcript information
Processing splicing information for gff3 file.
Deducing exon rank from relative coordinates provided
Prepare the 'metadata' data frame ... metadata: OK
Now generating chrominfo from available sequence names. No chromosome length information is available.
Warning messages:
1: In .local(con, format, text, ...) :
  gff-version directive indicates version is   3, not 3
2: In .deduceExonRankings(exs, format = "gff") :
  Infering Exon Rankings.  If this is not what you expected, then please be sure that you have provided a valid attribute for exonRankAttributeName
3: In matchCircularity(chroms, circ_seqs) :
  None of the strings in your circ_seqs argument match your seqnames.
4: 'dbBeginTransaction' is deprecated.
Use 'dbBegin' instead.
See help("Deprecated") 
5: 'dbBeginTransaction' is deprecated.
Use 'dbBegin' instead.
See help("Deprecated") 
6: 'dbBeginTransaction' is deprecated.
Use 'dbBegin' instead.
See help("Deprecated") 
7: 'dbBeginTransaction' is deprecated.
Use 'dbBegin' instead.
See help("Deprecated") 
8: 'dbBeginTransaction' is deprecated.
Use 'dbBegin' instead.
See help("Deprecated") 
9: 'dbBeginTransaction' is deprecated.
Use 'dbBegin' instead.
See help("Deprecated")

Agnieszka