lncRNA don't display correctly on GBrowse ?

classic Classic list List threaded Threaded
8 messages Options
Reply | Threaded
Open this post in threaded view
|

lncRNA don't display correctly on GBrowse ?

Bérénice Benayoun
Dear GBrowse developers,

I have been trying to display predicted lncRNA of my favorite species (a fish whose genome I am annotating). I am running Gbrowse out of an AWS machine using the public image as a starting point.

Everything is working well, except for lncRNAs: I can't get them to properly display the exon/intron structure (and the gene glyph doesn't look the same as for coding genes, through I configured everything identically save for the color). I attached a screen shot example...

I created a gff3 file derived from my Cufflinks results (I generated the gene, mRNA and exon features, similarly to my coding genes, but no CDS, since these are non-coding). I have loaded them using a SQLite backend.

Here is my invocation of the database in my conf file:

[lncRNA:database]
db_adaptor    = Bio::DB::SeqFeature::Store
db_args       = -adaptor DBI::SQLite
                -dsn     /opt/gbrowse/databases/NotFur/Nfur_lnc.sqlite
search options = default

Here is the track invocation:
[GenesNC]
database           = lncRNA
feature            = gene
glyph              = gene
#sub_part           = exon
height             = 15
bgcolor            = skyblue
connector_color    = midnightblue
decorate_introns   = 1
description        = 1
label              = 1
category           = Genes
key                = lncRNA genes
link               = AUTO

What am I doing wrong ? Is there any way to display non coding RNAs correctly ?

Thank you so much in advance for your help - I've spent hours on it without progress and this is driving me insane !!!

Best,

Berenice


--
Bérénice A. BENAYOUN, Ph.D.
Stanford University/Genetics Department
BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
M312 Alway Building
300, Pasteur Drive
MC 5120
Stanford, CA 94305-5120

USA
Email: [hidden email]
Web: www.stanford.edu/group/brunet/

------------------------------------------------------------------------------
Infragistics Professional
Build stunning WinForms apps today!
Reboot your WinForms applications with our WinForms controls.
Build a bridge from your legacy apps to the future.
http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
_______________________________________________
Gmod-gbrowse mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse

=?UTF-8?B?Q2FwdHVyZSBk4oCZZcyBY3JhbiAyMDE0LTA3LTI4IGHMgCA3LjQxLjAzIFBNLnBuZw==?= (64K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: lncRNA don't display correctly on GBrowse ?

Todd Harris
Hi Berenice -

What happens if you uncomment the “sub_part” option? That should force these entries to use exons as the subpart instead of CDS. Here’s an excerpt from the gene glyph documentation:

This glyph is designed to work properly with GFF3-style three-tier
genes, in which the top level feature has the Sequence Ontology type
of "gene", the second level feature(s) have the SO type "mRNA", and
the third level feature(s) have the SO type "CDS", "five_prime_utr"
and "three_prime_utr."  Subparts named "UTR" are also honored.  The
feature can contain other subparts as well (e.g. exon, intron,
translation), but they are currently ignored unless the option
sub_part is supplied.  If the sub_part option is used that feature
type will be used and CDS and UTR features will be excluded.
This could be used for specifying that exons be used instead,
for example.

Todd



On Jul 29, 2014, at 9:43 AM, Bérénice Benayoun <[hidden email]> wrote:

> Dear GBrowse developers,
>
> I have been trying to display predicted lncRNA of my favorite species (a fish whose genome I am annotating). I am running Gbrowse out of an AWS machine using the public image as a starting point.
>
> Everything is working well, except for lncRNAs: I can't get them to properly display the exon/intron structure (and the gene glyph doesn't look the same as for coding genes, through I configured everything identically save for the color). I attached a screen shot example...
>
> I created a gff3 file derived from my Cufflinks results (I generated the gene, mRNA and exon features, similarly to my coding genes, but no CDS, since these are non-coding). I have loaded them using a SQLite backend.
>
> Here is my invocation of the database in my conf file:
>
> [lncRNA:database]
> db_adaptor    = Bio::DB::SeqFeature::Store
> db_args       = -adaptor DBI::SQLite
>                 -dsn     /opt/gbrowse/databases/NotFur/Nfur_lnc.sqlite
> search options = default
>
> Here is the track invocation:
> [GenesNC]
> database           = lncRNA
> feature            = gene
> glyph              = gene
> #sub_part           = exon
> height             = 15
> bgcolor            = skyblue
> connector_color    = midnightblue
> decorate_introns   = 1
> description        = 1
> label              = 1
> category           = Genes
> key                = lncRNA genes
> link               = AUTO
>
> What am I doing wrong ? Is there any way to display non coding RNAs correctly ?
>
> Thank you so much in advance for your help - I've spent hours on it without progress and this is driving me insane !!!
>
> Best,
>
> Berenice
>
>
> --
> Bérénice A. BENAYOUN, Ph.D.
> Stanford University/Genetics Department
> BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> M312 Alway Building
> 300, Pasteur Drive
> MC 5120
> Stanford, CA 94305-5120
> USA
> Email: [hidden email]
> Web: www.stanford.edu/group/brunet/
> <Capture d’écran 2014-07-28 à 7.41.03 PM.png>------------------------------------------------------------------------------
> Infragistics Professional
> Build stunning WinForms apps today!
> Reboot your WinForms applications with our WinForms controls.
> Build a bridge from your legacy apps to the future.
> http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> Gmod-gbrowse mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse


------------------------------------------------------------------------------
Infragistics Professional
Build stunning WinForms apps today!
Reboot your WinForms applications with our WinForms controls.
Build a bridge from your legacy apps to the future.
http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
_______________________________________________
Gmod-gbrowse mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
Reply | Threaded
Open this post in threaded view
|

Re: lncRNA don't display correctly on GBrowse ?

Bérénice Benayoun
Hi Todd,

thanks for the quick answer !

Actually, this was a previous try when I was trying to get this to work, and I commented it out to see if it made any difference... I de-commented it again, but no change, the multi-exon lncRNA is still one block...

is there any hack or any other glyph that I could use that could at least show the exons connected but distinct ? I tried segments, but it also gave me a big block (though there were some vertical lines in there)...

Thank you so much for your help, I really appreciate it !

best,

Berenice


2014-07-28 21:19 GMT-07:00 Todd Harris <[hidden email]>:
Hi Berenice -

What happens if you uncomment the “sub_part” option? That should force these entries to use exons as the subpart instead of CDS. Here’s an excerpt from the gene glyph documentation:

This glyph is designed to work properly with GFF3-style three-tier
genes, in which the top level feature has the Sequence Ontology type
of "gene", the second level feature(s) have the SO type "mRNA", and
the third level feature(s) have the SO type "CDS", "five_prime_utr"
and "three_prime_utr."  Subparts named "UTR" are also honored.  The
feature can contain other subparts as well (e.g. exon, intron,
translation), but they are currently ignored unless the option
sub_part is supplied.  If the sub_part option is used that feature
type will be used and CDS and UTR features will be excluded.
This could be used for specifying that exons be used instead,
for example.

Todd



On Jul 29, 2014, at 9:43 AM, Bérénice Benayoun <[hidden email]> wrote:

> Dear GBrowse developers,
>
> I have been trying to display predicted lncRNA of my favorite species (a fish whose genome I am annotating). I am running Gbrowse out of an AWS machine using the public image as a starting point.
>
> Everything is working well, except for lncRNAs: I can't get them to properly display the exon/intron structure (and the gene glyph doesn't look the same as for coding genes, through I configured everything identically save for the color). I attached a screen shot example...
>
> I created a gff3 file derived from my Cufflinks results (I generated the gene, mRNA and exon features, similarly to my coding genes, but no CDS, since these are non-coding). I have loaded them using a SQLite backend.
>
> Here is my invocation of the database in my conf file:
>
> [lncRNA:database]
> db_adaptor    = Bio::DB::SeqFeature::Store
> db_args       = -adaptor DBI::SQLite
>                 -dsn     /opt/gbrowse/databases/NotFur/Nfur_lnc.sqlite
> search options = default
>
> Here is the track invocation:
> [GenesNC]
> database           = lncRNA
> feature            = gene
> glyph              = gene
> #sub_part           = exon
> height             = 15
> bgcolor            = skyblue
> connector_color    = midnightblue
> decorate_introns   = 1
> description        = 1
> label              = 1
> category           = Genes
> key                = lncRNA genes
> link               = AUTO
>
> What am I doing wrong ? Is there any way to display non coding RNAs correctly ?
>
> Thank you so much in advance for your help - I've spent hours on it without progress and this is driving me insane !!!
>
> Best,
>
> Berenice
>
>
> --
> Bérénice A. BENAYOUN, Ph.D.
> Stanford University/Genetics Department
> BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> M312 Alway Building
> 300, Pasteur Drive
> MC 5120
> Stanford, CA 94305-5120
> USA
> Email: [hidden email]
> Web: www.stanford.edu/group/brunet/
> <Capture d’écran 2014-07-28 à 7.41.03 PM.png>------------------------------------------------------------------------------
> Infragistics Professional
> Build stunning WinForms apps today!
> Reboot your WinForms applications with our WinForms controls.
> Build a bridge from your legacy apps to the future.
> http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> Gmod-gbrowse mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse




--
Bérénice A. BENAYOUN, Ph.D.
Stanford University/Genetics Department
BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
M312 Alway Building
300, Pasteur Drive
MC 5120
Stanford, CA 94305-5120

USA
Email: [hidden email]
Web: www.stanford.edu/group/brunet/

------------------------------------------------------------------------------
Infragistics Professional
Build stunning WinForms apps today!
Reboot your WinForms applications with our WinForms controls.
Build a bridge from your legacy apps to the future.
http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
_______________________________________________
Gmod-gbrowse mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
Reply | Threaded
Open this post in threaded view
|

Re: lncRNA don't display correctly on GBrowse ?

Michael Dondrup-3
Hi,

I think this is due to the way the Bio::Graphics::Glyph::gene glyph works, you can have a look at its perldoc at http://search.cpan.org/~lds/Bio-Graphics/lib/Bio/Graphics/Glyph/gene.pm.
The glyph works correctly only in case the Sequence ontology terms follow the exact hierarchy "gene"->"mRNA"->"exon"->"three_prime_utr"/ "five_prime_utr".
The glyph doesn't resolve the SO hierarchy to find if e.g. rna_gene or lncrna_gene is also derived from gene. If it doesn't understand the type, it defaults to the
'standard' glyph which is just a block. There are two possibilities I can think of:

- edit your GFF file, replace all SO terms by the ones that Bio::Graphics::Glyph::gene uses and update your features
- derive your own glyph class from the source code of Bio::Graphics::Glyph::gene e.g. Bio::Graphics::Glyph::lncrna_gene and
replace all occurrences  of SO terms in the code, most or all are found in regular expressions.

Hope this helps
Michael

Michael Dondrup
Postdoctoral fellow
Sea Lice Research Centre/Department of Informatics
University of Bergen
Thormøhlensgate 55, N-5008 Bergen,
Norway

On Jul 29, 2014, at 7:20 PM, Bérénice Benayoun wrote:

> Hi Todd,
>
> thanks for the quick answer !
>
> Actually, this was a previous try when I was trying to get this to work, and I commented it out to see if it made any difference... I de-commented it again, but no change, the multi-exon lncRNA is still one block...
>
> is there any hack or any other glyph that I could use that could at least show the exons connected but distinct ? I tried segments, but it also gave me a big block (though there were some vertical lines in there)...
>
> Thank you so much for your help, I really appreciate it !
>
> best,
>
> Berenice
>
>
> 2014-07-28 21:19 GMT-07:00 Todd Harris <[hidden email]>:
> Hi Berenice -
>
> What happens if you uncomment the “sub_part” option? That should force these entries to use exons as the subpart instead of CDS. Here’s an excerpt from the gene glyph documentation:
>
> This glyph is designed to work properly with GFF3-style three-tier
> genes, in which the top level feature has the Sequence Ontology type
> of "gene", the second level feature(s) have the SO type "mRNA", and
> the third level feature(s) have the SO type "CDS", "five_prime_utr"
> and "three_prime_utr."  Subparts named "UTR" are also honored.  The
> feature can contain other subparts as well (e.g. exon, intron,
> translation), but they are currently ignored unless the option
> sub_part is supplied.  If the sub_part option is used that feature
> type will be used and CDS and UTR features will be excluded.
> This could be used for specifying that exons be used instead,
> for example.
>
> Todd
>
>
>
> On Jul 29, 2014, at 9:43 AM, Bérénice Benayoun <[hidden email]> wrote:
>
> > Dear GBrowse developers,
> >
> > I have been trying to display predicted lncRNA of my favorite species (a fish whose genome I am annotating). I am running Gbrowse out of an AWS machine using the public image as a starting point.
> >
> > Everything is working well, except for lncRNAs: I can't get them to properly display the exon/intron structure (and the gene glyph doesn't look the same as for coding genes, through I configured everything identically save for the color). I attached a screen shot example...
> >
> > I created a gff3 file derived from my Cufflinks results (I generated the gene, mRNA and exon features, similarly to my coding genes, but no CDS, since these are non-coding). I have loaded them using a SQLite backend.
> >
> > Here is my invocation of the database in my conf file:
> >
> > [lncRNA:database]
> > db_adaptor    = Bio::DB::SeqFeature::Store
> > db_args       = -adaptor DBI::SQLite
> >                 -dsn     /opt/gbrowse/databases/NotFur/Nfur_lnc.sqlite
> > search options = default
> >
> > Here is the track invocation:
> > [GenesNC]
> > database           = lncRNA
> > feature            = gene
> > glyph              = gene
> > #sub_part           = exon
> > height             = 15
> > bgcolor            = skyblue
> > connector_color    = midnightblue
> > decorate_introns   = 1
> > description        = 1
> > label              = 1
> > category           = Genes
> > key                = lncRNA genes
> > link               = AUTO
> >
> > What am I doing wrong ? Is there any way to display non coding RNAs correctly ?
> >
> > Thank you so much in advance for your help - I've spent hours on it without progress and this is driving me insane !!!
> >
> > Best,
> >
> > Berenice
> >
> >
> > --
> > Bérénice A. BENAYOUN, Ph.D.
> > Stanford University/Genetics Department
> > BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> > M312 Alway Building
> > 300, Pasteur Drive
> > MC 5120
> > Stanford, CA 94305-5120
> > USA
> > Email: [hidden email]
> > Web: www.stanford.edu/group/brunet/
> > <Capture d’écran 2014-07-28 à 7.41.03 PM.png>------------------------------------------------------------------------------
> > Infragistics Professional
> > Build stunning WinForms apps today!
> > Reboot your WinForms applications with our WinForms controls.
> > Build a bridge from your legacy apps to the future.
> > http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> > Gmod-gbrowse mailing list
> > [hidden email]
> > https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
>
>
>
>
> --
> Bérénice A. BENAYOUN, Ph.D.
> Stanford University/Genetics Department
> BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> M312 Alway Building
> 300, Pasteur Drive
> MC 5120
> Stanford, CA 94305-5120
> USA
> Email: [hidden email]
> Web: www.stanford.edu/group/brunet/
> ------------------------------------------------------------------------------
> Infragistics Professional
> Build stunning WinForms apps today!
> Reboot your WinForms applications with our WinForms controls.
> Build a bridge from your legacy apps to the future.
> http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> Gmod-gbrowse mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse

------------------------------------------------------------------------------
Infragistics Professional
Build stunning WinForms apps today!
Reboot your WinForms applications with our WinForms controls.
Build a bridge from your legacy apps to the future.
http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
_______________________________________________
Gmod-gbrowse mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse

signature.asc (465 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: lncRNA don't display correctly on GBrowse ?

Scott Cain
Hi Berenice,

It should really work.  The sample of GFF below is from WormBase (I trimmed it a little to make it easier to read).

III  .  gene    832984  833347  .   +   .   ID=Gene:WBGene00219709
III  .   lincRNA 832984  833347  .   +   .       ID=Transcript:K02F3.14;Parent=Gene:WBGene00219709
III   .  exon    832984  833061  .   +   .   Parent=Transcript:K02F3.14
III   .  exon    833124  833347  .   +   .   Parent=Transcript:K02F3.14

I assume yours looks similar, right?  This lincRNA gene is represented in GBrowse in both the Curated Genes and Curated Genes (noncoding) tracks:


The coding genes in the Curated Genes track are gene->mRNA->CDS/UTR/exon (but it's clearly using the CDS/UTR data to draw the glyphs since it's coloring the UTRs appropriately).  There's nothing special about the track config, though it does have box_subparts turned on--I'm not sure what it's doing in this context.

Scott





On Tue, Jul 29, 2014 at 1:50 PM, Michael Dondrup <[hidden email]> wrote:
>
> Hi,
>
> I think this is due to the way the Bio::Graphics::Glyph::gene glyph works, you can have a look at its perldoc at http://search.cpan.org/~lds/Bio-Graphics/lib/Bio/Graphics/Glyph/gene.pm.
> The glyph works correctly only in case the Sequence ontology terms follow the exact hierarchy "gene"->"mRNA"->"exon"->"three_prime_utr"/ "five_prime_utr".
> The glyph doesn't resolve the SO hierarchy to find if e.g. rna_gene or lncrna_gene is also derived from gene. If it doesn't understand the type, it defaults to the
> 'standard' glyph which is just a block. There are two possibilities I can think of:
>
> - edit your GFF file, replace all SO terms by the ones that Bio::Graphics::Glyph::gene uses and update your features
> - derive your own glyph class from the source code of Bio::Graphics::Glyph::gene e.g. Bio::Graphics::Glyph::lncrna_gene and
> replace all occurrences  of SO terms in the code, most or all are found in regular expressions.
>
> Hope this helps
> Michael
>
> Michael Dondrup
> Postdoctoral fellow
> Sea Lice Research Centre/Department of Informatics
> University of Bergen
> Thormøhlensgate 55, N-5008 Bergen,
> Norway
>
> On Jul 29, 2014, at 7:20 PM, Bérénice Benayoun wrote:
>
> > Hi Todd,
> >
> > thanks for the quick answer !
> >
> > Actually, this was a previous try when I was trying to get this to work, and I commented it out to see if it made any difference... I de-commented it again, but no change, the multi-exon lncRNA is still one block...
> >
> > is there any hack or any other glyph that I could use that could at least show the exons connected but distinct ? I tried segments, but it also gave me a big block (though there were some vertical lines in there)...
> >
> > Thank you so much for your help, I really appreciate it !
> >
> > best,
> >
> > Berenice
> >
> >
> > 2014-07-28 21:19 GMT-07:00 Todd Harris <[hidden email]>:
> > Hi Berenice -
> >
> > What happens if you uncomment the “sub_part” option? That should force these entries to use exons as the subpart instead of CDS. Here’s an excerpt from the gene glyph documentation:
> >
> > This glyph is designed to work properly with GFF3-style three-tier
> > genes, in which the top level feature has the Sequence Ontology type
> > of "gene", the second level feature(s) have the SO type "mRNA", and
> > the third level feature(s) have the SO type "CDS", "five_prime_utr"
> > and "three_prime_utr."  Subparts named "UTR" are also honored.  The
> > feature can contain other subparts as well (e.g. exon, intron,
> > translation), but they are currently ignored unless the option
> > sub_part is supplied.  If the sub_part option is used that feature
> > type will be used and CDS and UTR features will be excluded.
> > This could be used for specifying that exons be used instead,
> > for example.
> >
> > Todd
> >
> >
> >
> > On Jul 29, 2014, at 9:43 AM, Bérénice Benayoun <[hidden email]> wrote:
> >
> > > Dear GBrowse developers,
> > >
> > > I have been trying to display predicted lncRNA of my favorite species (a fish whose genome I am annotating). I am running Gbrowse out of an AWS machine using the public image as a starting point.
> > >
> > > Everything is working well, except for lncRNAs: I can't get them to properly display the exon/intron structure (and the gene glyph doesn't look the same as for coding genes, through I configured everything identically save for the color). I attached a screen shot example...
> > >
> > > I created a gff3 file derived from my Cufflinks results (I generated the gene, mRNA and exon features, similarly to my coding genes, but no CDS, since these are non-coding). I have loaded them using a SQLite backend.
> > >
> > > Here is my invocation of the database in my conf file:
> > >
> > > [lncRNA:database]
> > > db_adaptor    = Bio::DB::SeqFeature::Store
> > > db_args       = -adaptor DBI::SQLite
> > >                 -dsn     /opt/gbrowse/databases/NotFur/Nfur_lnc.sqlite
> > > search options = default
> > >
> > > Here is the track invocation:
> > > [GenesNC]
> > > database           = lncRNA
> > > feature            = gene
> > > glyph              = gene
> > > #sub_part           = exon
> > > height             = 15
> > > bgcolor            = skyblue
> > > connector_color    = midnightblue
> > > decorate_introns   = 1
> > > description        = 1
> > > label              = 1
> > > category           = Genes
> > > key                = lncRNA genes
> > > link               = AUTO
> > >
> > > What am I doing wrong ? Is there any way to display non coding RNAs correctly ?
> > >
> > > Thank you so much in advance for your help - I've spent hours on it without progress and this is driving me insane !!!
> > >
> > > Best,
> > >
> > > Berenice
> > >
> > >
> > > --
> > > Bérénice A. BENAYOUN, Ph.D.
> > > Stanford University/Genetics Department
> > > BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> > > M312 Alway Building
> > > 300, Pasteur Drive
> > > MC 5120
> > > Stanford, CA 94305-5120
> > > USA
> > > Email: [hidden email]
> > > Web: www.stanford.edu/group/brunet/
> > > <Capture d’écran 2014-07-28 à 7.41.03 PM.png>------------------------------------------------------------------------------
> > > Infragistics Professional
> > > Build stunning WinForms apps today!
> > > Reboot your WinForms applications with our WinForms controls.
> > > Build a bridge from your legacy apps to the future.
> > > http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> > > Gmod-gbrowse mailing list
> > > [hidden email]
> > > https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
> >
> >
> >
> >
> > --
> > Bérénice A. BENAYOUN, Ph.D.
> > Stanford University/Genetics Department
> > BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> > M312 Alway Building
> > 300, Pasteur Drive
> > MC 5120
> > Stanford, CA 94305-5120
> > USA
> > Email: [hidden email]
> > Web: www.stanford.edu/group/brunet/
> > ------------------------------------------------------------------------------
> > Infragistics Professional
> > Build stunning WinForms apps today!
> > Reboot your WinForms applications with our WinForms controls.
> > Build a bridge from your legacy apps to the future.
> > http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> > Gmod-gbrowse mailing list
> > [hidden email]
> > https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
>
>
> ------------------------------------------------------------------------------
> Infragistics Professional
> Build stunning WinForms apps today!
> Reboot your WinForms applications with our WinForms controls.
> Build a bridge from your legacy apps to the future.
> http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
> _______________________________________________
> Gmod-gbrowse mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
>



--
------------------------------------------------------------------------
Scott Cain, Ph. D.                                   scott at scottcain dot net
GMOD Coordinator (http://gmod.org/)                     216-392-3087
Ontario Institute for Cancer Research

------------------------------------------------------------------------------
Infragistics Professional
Build stunning WinForms apps today!
Reboot your WinForms applications with our WinForms controls.
Build a bridge from your legacy apps to the future.
http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
_______________________________________________
Gmod-gbrowse mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
Reply | Threaded
Open this post in threaded view
|

Re: lncRNA don't display correctly on GBrowse ?

Bérénice Benayoun
Hi Scott and Michael,

Unfortunately, writing a derived class is beyond me (I look at the existing ones and I am unable to understand what they do without more comments). I tried doing just the UTR/exon substitution to make a lincRNA class, but I only got a red band with an error message in the browser.

The problem is that I also want to avoid putting CDSs where it doesn't make biological sense (there are obviously non coding...).

I had seen such tracks in UCSC and Wormbase (such as sent by Scott), so I thought it should be such a problem to get them to work in a new GBrowse derived browser.

Here is an extract of my non-coding GFF:

GapFilledScaffold_1264 Cufflinks gene 64907 65350 1000 - . ID=979895;Name=NFURLNR01264041000;Note=Cufflinks predicted non-coding
GapFilledScaffold_1264 Cufflinks mRNA 64907 65350 1000 - . ID=979896;Name=NFURLNR01264041000-mRNA-1;Parent=979895;Note=Cufflinks predicted non-coding
GapFilledScaffold_1264 Cufflinks exon 64907 64931 1000 - . ID=979897;Name=NFURLNR01264041000-mRNA-1:exon:1;Parent=979895;Note=Cufflinks predicted non-coding
GapFilledScaffold_1264 Cufflinks exon 65221 65350 1000 - . ID=979898;Name=NFURLNR01264041000-mRNA-1:exon:2;Parent=979895;Note=Cufflinks predicted non-coding


It has the parents relationships that should explain inheritance/parts to the glyph?

Maybe Wormbase has derived their own glyphs for non-coding ? Otherwise maybe my configurations are wrong ? Do you know if I can put my hands on the track configuration they use to see how it differs from mine ?

I looked through the code of processed_transcript and it seems to also require CDSs. With trasncript2, the best I can get is just vertical separation (see attached screencap) but the introns are displayed same as exons.

Sorry for my beginner questions - I am just really at a loss...

Thank you so much for your help !!!

Berenice



2014-07-29 12:03 GMT-07:00 Scott Cain <[hidden email]>:
Hi Berenice,

It should really work.  The sample of GFF below is from WormBase (I trimmed it a little to make it easier to read).

III  .  gene    832984  833347  .   +   .   ID=Gene:WBGene00219709
III  .   lincRNA 832984  833347  .   +   .       ID=Transcript:K02F3.14;Parent=Gene:WBGene00219709
III   .  exon    832984  833061  .   +   .   Parent=Transcript:K02F3.14
III   .  exon    833124  833347  .   +   .   Parent=Transcript:K02F3.14

I assume yours looks similar, right?  This lincRNA gene is represented in GBrowse in both the Curated Genes and Curated Genes (noncoding) tracks:


The coding genes in the Curated Genes track are gene->mRNA->CDS/UTR/exon (but it's clearly using the CDS/UTR data to draw the glyphs since it's coloring the UTRs appropriately).  There's nothing special about the track config, though it does have box_subparts turned on--I'm not sure what it's doing in this context.

Scott





On Tue, Jul 29, 2014 at 1:50 PM, Michael Dondrup <[hidden email]> wrote:
>
> Hi,
>
> I think this is due to the way the Bio::Graphics::Glyph::gene glyph works, you can have a look at its perldoc at http://search.cpan.org/~lds/Bio-Graphics/lib/Bio/Graphics/Glyph/gene.pm.
> The glyph works correctly only in case the Sequence ontology terms follow the exact hierarchy "gene"->"mRNA"->"exon"->"three_prime_utr"/ "five_prime_utr".
> The glyph doesn't resolve the SO hierarchy to find if e.g. rna_gene or lncrna_gene is also derived from gene. If it doesn't understand the type, it defaults to the
> 'standard' glyph which is just a block. There are two possibilities I can think of:
>
> - edit your GFF file, replace all SO terms by the ones that Bio::Graphics::Glyph::gene uses and update your features
> - derive your own glyph class from the source code of Bio::Graphics::Glyph::gene e.g. Bio::Graphics::Glyph::lncrna_gene and
> replace all occurrences  of SO terms in the code, most or all are found in regular expressions.
>
> Hope this helps
> Michael
>
> Michael Dondrup
> Postdoctoral fellow
> Sea Lice Research Centre/Department of Informatics
> University of Bergen
> Thormøhlensgate 55, N-5008 Bergen,
> Norway
>
> On Jul 29, 2014, at 7:20 PM, Bérénice Benayoun wrote:
>
> > Hi Todd,
> >
> > thanks for the quick answer !
> >
> > Actually, this was a previous try when I was trying to get this to work, and I commented it out to see if it made any difference... I de-commented it again, but no change, the multi-exon lncRNA is still one block...
> >
> > is there any hack or any other glyph that I could use that could at least show the exons connected but distinct ? I tried segments, but it also gave me a big block (though there were some vertical lines in there)...
> >
> > Thank you so much for your help, I really appreciate it !
> >
> > best,
> >
> > Berenice
> >
> >
> > 2014-07-28 21:19 GMT-07:00 Todd Harris <[hidden email]>:
> > Hi Berenice -
> >
> > What happens if you uncomment the “sub_part” option? That should force these entries to use exons as the subpart instead of CDS. Here’s an excerpt from the gene glyph documentation:
> >
> > This glyph is designed to work properly with GFF3-style three-tier
> > genes, in which the top level feature has the Sequence Ontology type
> > of "gene", the second level feature(s) have the SO type "mRNA", and
> > the third level feature(s) have the SO type "CDS", "five_prime_utr"
> > and "three_prime_utr."  Subparts named "UTR" are also honored.  The
> > feature can contain other subparts as well (e.g. exon, intron,
> > translation), but they are currently ignored unless the option
> > sub_part is supplied.  If the sub_part option is used that feature
> > type will be used and CDS and UTR features will be excluded.
> > This could be used for specifying that exons be used instead,
> > for example.
> >
> > Todd
> >
> >
> >
> > On Jul 29, 2014, at 9:43 AM, Bérénice Benayoun <[hidden email]> wrote:
> >
> > > Dear GBrowse developers,
> > >
> > > I have been trying to display predicted lncRNA of my favorite species (a fish whose genome I am annotating). I am running Gbrowse out of an AWS machine using the public image as a starting point.
> > >
> > > Everything is working well, except for lncRNAs: I can't get them to properly display the exon/intron structure (and the gene glyph doesn't look the same as for coding genes, through I configured everything identically save for the color). I attached a screen shot example...
> > >
> > > I created a gff3 file derived from my Cufflinks results (I generated the gene, mRNA and exon features, similarly to my coding genes, but no CDS, since these are non-coding). I have loaded them using a SQLite backend.
> > >
> > > Here is my invocation of the database in my conf file:
> > >
> > > [lncRNA:database]
> > > db_adaptor    = Bio::DB::SeqFeature::Store
> > > db_args       = -adaptor DBI::SQLite
> > >                 -dsn     /opt/gbrowse/databases/NotFur/Nfur_lnc.sqlite
> > > search options = default
> > >
> > > Here is the track invocation:
> > > [GenesNC]
> > > database           = lncRNA
> > > feature            = gene
> > > glyph              = gene
> > > #sub_part           = exon
> > > height             = 15
> > > bgcolor            = skyblue
> > > connector_color    = midnightblue
> > > decorate_introns   = 1
> > > description        = 1
> > > label              = 1
> > > category           = Genes
> > > key                = lncRNA genes
> > > link               = AUTO
> > >
> > > What am I doing wrong ? Is there any way to display non coding RNAs correctly ?
> > >
> > > Thank you so much in advance for your help - I've spent hours on it without progress and this is driving me insane !!!
> > >
> > > Best,
> > >
> > > Berenice
> > >
> > >
> > > --
> > > Bérénice A. BENAYOUN, Ph.D.
> > > Stanford University/Genetics Department
> > > BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> > > M312 Alway Building
> > > 300, Pasteur Drive
> > > MC 5120
> > > Stanford, CA 94305-5120
> > > USA
> > > Email: [hidden email]
> > > Web: www.stanford.edu/group/brunet/
> > > <Capture d’écran 2014-07-28 à 7.41.03 PM.png>------------------------------------------------------------------------------
> > > Infragistics Professional
> > > Build stunning WinForms apps today!
> > > Reboot your WinForms applications with our WinForms controls.
> > > Build a bridge from your legacy apps to the future.
> > > http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> > > Gmod-gbrowse mailing list
> > > [hidden email]
> > > https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
> >
> >
> >
> >
> > --
> > Bérénice A. BENAYOUN, Ph.D.
> > Stanford University/Genetics Department
> > BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> > M312 Alway Building
> > 300, Pasteur Drive
> > MC 5120
> > Stanford, CA 94305-5120
> > USA
> > Email: [hidden email]
> > Web: www.stanford.edu/group/brunet/
> > ------------------------------------------------------------------------------
> > Infragistics Professional
> > Build stunning WinForms apps today!
> > Reboot your WinForms applications with our WinForms controls.
> > Build a bridge from your legacy apps to the future.
> > http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> > Gmod-gbrowse mailing list
> > [hidden email]
> > https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
>
>
> ------------------------------------------------------------------------------
> Infragistics Professional
> Build stunning WinForms apps today!
> Reboot your WinForms applications with our WinForms controls.
> Build a bridge from your legacy apps to the future.
> http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
> _______________________________________________
> Gmod-gbrowse mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
>



--
------------------------------------------------------------------------
Scott Cain, Ph. D.                                   scott at scottcain dot net
GMOD Coordinator (http://gmod.org/)                     <a href="tel:216-392-3087" value="+12163923087" target="_blank">216-392-3087
Ontario Institute for Cancer Research



--
Bérénice A. BENAYOUN, Ph.D.
Stanford University/Genetics Department
BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
M312 Alway Building
300, Pasteur Drive
MC 5120
Stanford, CA 94305-5120

USA
Email: [hidden email]
Web: www.stanford.edu/group/brunet/

------------------------------------------------------------------------------
Infragistics Professional
Build stunning WinForms apps today!
Reboot your WinForms applications with our WinForms controls.
Build a bridge from your legacy apps to the future.
http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
_______________________________________________
Gmod-gbrowse mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse

=?UTF-8?B?Q2FwdHVyZSBk4oCZZcyBY3JhbiAyMDE0LTA3LTI5IGHMgCAxLjM3LjM0IFBNLnBuZw==?= (62K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: lncRNA don't display correctly on GBrowse ?

Scott Cain
Hi Berenice,

WormBase is just using the standard gene glyph.  I think the problem is with your GFF: it has the parent of the exons being the gene, whereas the parent should be the mRNA (which is also a bit of a problem--clearly these aren't mRNAs either but that is a minor detail).  If you had explicit introns in the gff, I would suggest adding the "ignore_sub_part = intron" option, but it doesn't look like that's the case.

Scott



On Tue, Jul 29, 2014 at 4:41 PM, Bérénice Benayoun <[hidden email]> wrote:
Hi Scott and Michael,

Unfortunately, writing a derived class is beyond me (I look at the existing ones and I am unable to understand what they do without more comments). I tried doing just the UTR/exon substitution to make a lincRNA class, but I only got a red band with an error message in the browser.

The problem is that I also want to avoid putting CDSs where it doesn't make biological sense (there are obviously non coding...).

I had seen such tracks in UCSC and Wormbase (such as sent by Scott), so I thought it should be such a problem to get them to work in a new GBrowse derived browser.

Here is an extract of my non-coding GFF:

GapFilledScaffold_1264 Cufflinks gene 64907 65350 1000 - . ID=979895;Name=NFURLNR01264041000;Note=Cufflinks predicted non-coding
GapFilledScaffold_1264 Cufflinks mRNA 64907 65350 1000 - . ID=979896;Name=NFURLNR01264041000-mRNA-1;Parent=979895;Note=Cufflinks predicted non-coding
GapFilledScaffold_1264 Cufflinks exon 64907 64931 1000 - . ID=979897;Name=NFURLNR01264041000-mRNA-1:exon:1;Parent=979895;Note=Cufflinks predicted non-coding
GapFilledScaffold_1264 Cufflinks exon 65221 65350 1000 - . ID=979898;Name=NFURLNR01264041000-mRNA-1:exon:2;Parent=979895;Note=Cufflinks predicted non-coding


It has the parents relationships that should explain inheritance/parts to the glyph?

Maybe Wormbase has derived their own glyphs for non-coding ? Otherwise maybe my configurations are wrong ? Do you know if I can put my hands on the track configuration they use to see how it differs from mine ?

I looked through the code of processed_transcript and it seems to also require CDSs. With trasncript2, the best I can get is just vertical separation (see attached screencap) but the introns are displayed same as exons.

Sorry for my beginner questions - I am just really at a loss...

Thank you so much for your help !!!

Berenice



2014-07-29 12:03 GMT-07:00 Scott Cain <[hidden email]>:

Hi Berenice,

It should really work.  The sample of GFF below is from WormBase (I trimmed it a little to make it easier to read).

III  .  gene    832984  833347  .   +   .   ID=Gene:WBGene00219709
III  .   lincRNA 832984  833347  .   +   .       ID=Transcript:K02F3.14;Parent=Gene:WBGene00219709
III   .  exon    832984  833061  .   +   .   Parent=Transcript:K02F3.14
III   .  exon    833124  833347  .   +   .   Parent=Transcript:K02F3.14

I assume yours looks similar, right?  This lincRNA gene is represented in GBrowse in both the Curated Genes and Curated Genes (noncoding) tracks:


The coding genes in the Curated Genes track are gene->mRNA->CDS/UTR/exon (but it's clearly using the CDS/UTR data to draw the glyphs since it's coloring the UTRs appropriately).  There's nothing special about the track config, though it does have box_subparts turned on--I'm not sure what it's doing in this context.

Scott





On Tue, Jul 29, 2014 at 1:50 PM, Michael Dondrup <[hidden email]> wrote:
>
> Hi,
>
> I think this is due to the way the Bio::Graphics::Glyph::gene glyph works, you can have a look at its perldoc at http://search.cpan.org/~lds/Bio-Graphics/lib/Bio/Graphics/Glyph/gene.pm.
> The glyph works correctly only in case the Sequence ontology terms follow the exact hierarchy "gene"->"mRNA"->"exon"->"three_prime_utr"/ "five_prime_utr".
> The glyph doesn't resolve the SO hierarchy to find if e.g. rna_gene or lncrna_gene is also derived from gene. If it doesn't understand the type, it defaults to the
> 'standard' glyph which is just a block. There are two possibilities I can think of:
>
> - edit your GFF file, replace all SO terms by the ones that Bio::Graphics::Glyph::gene uses and update your features
> - derive your own glyph class from the source code of Bio::Graphics::Glyph::gene e.g. Bio::Graphics::Glyph::lncrna_gene and
> replace all occurrences  of SO terms in the code, most or all are found in regular expressions.
>
> Hope this helps
> Michael
>
> Michael Dondrup
> Postdoctoral fellow
> Sea Lice Research Centre/Department of Informatics
> University of Bergen
> Thormøhlensgate 55, N-5008 Bergen,
> Norway
>
> On Jul 29, 2014, at 7:20 PM, Bérénice Benayoun wrote:
>
> > Hi Todd,
> >
> > thanks for the quick answer !
> >
> > Actually, this was a previous try when I was trying to get this to work, and I commented it out to see if it made any difference... I de-commented it again, but no change, the multi-exon lncRNA is still one block...
> >
> > is there any hack or any other glyph that I could use that could at least show the exons connected but distinct ? I tried segments, but it also gave me a big block (though there were some vertical lines in there)...
> >
> > Thank you so much for your help, I really appreciate it !
> >
> > best,
> >
> > Berenice
> >
> >
> > 2014-07-28 21:19 GMT-07:00 Todd Harris <[hidden email]>:
> > Hi Berenice -
> >
> > What happens if you uncomment the “sub_part” option? That should force these entries to use exons as the subpart instead of CDS. Here’s an excerpt from the gene glyph documentation:
> >
> > This glyph is designed to work properly with GFF3-style three-tier
> > genes, in which the top level feature has the Sequence Ontology type
> > of "gene", the second level feature(s) have the SO type "mRNA", and
> > the third level feature(s) have the SO type "CDS", "five_prime_utr"
> > and "three_prime_utr."  Subparts named "UTR" are also honored.  The
> > feature can contain other subparts as well (e.g. exon, intron,
> > translation), but they are currently ignored unless the option
> > sub_part is supplied.  If the sub_part option is used that feature
> > type will be used and CDS and UTR features will be excluded.
> > This could be used for specifying that exons be used instead,
> > for example.
> >
> > Todd
> >
> >
> >
> > On Jul 29, 2014, at 9:43 AM, Bérénice Benayoun <[hidden email]> wrote:
> >
> > > Dear GBrowse developers,
> > >
> > > I have been trying to display predicted lncRNA of my favorite species (a fish whose genome I am annotating). I am running Gbrowse out of an AWS machine using the public image as a starting point.
> > >
> > > Everything is working well, except for lncRNAs: I can't get them to properly display the exon/intron structure (and the gene glyph doesn't look the same as for coding genes, through I configured everything identically save for the color). I attached a screen shot example...
> > >
> > > I created a gff3 file derived from my Cufflinks results (I generated the gene, mRNA and exon features, similarly to my coding genes, but no CDS, since these are non-coding). I have loaded them using a SQLite backend.
> > >
> > > Here is my invocation of the database in my conf file:
> > >
> > > [lncRNA:database]
> > > db_adaptor    = Bio::DB::SeqFeature::Store
> > > db_args       = -adaptor DBI::SQLite
> > >                 -dsn     /opt/gbrowse/databases/NotFur/Nfur_lnc.sqlite
> > > search options = default
> > >
> > > Here is the track invocation:
> > > [GenesNC]
> > > database           = lncRNA
> > > feature            = gene
> > > glyph              = gene
> > > #sub_part           = exon
> > > height             = 15
> > > bgcolor            = skyblue
> > > connector_color    = midnightblue
> > > decorate_introns   = 1
> > > description        = 1
> > > label              = 1
> > > category           = Genes
> > > key                = lncRNA genes
> > > link               = AUTO
> > >
> > > What am I doing wrong ? Is there any way to display non coding RNAs correctly ?
> > >
> > > Thank you so much in advance for your help - I've spent hours on it without progress and this is driving me insane !!!
> > >
> > > Best,
> > >
> > > Berenice
> > >
> > >
> > > --
> > > Bérénice A. BENAYOUN, Ph.D.
> > > Stanford University/Genetics Department
> > > BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> > > M312 Alway Building
> > > 300, Pasteur Drive
> > > MC 5120
> > > Stanford, CA 94305-5120
> > > USA
> > > Email: [hidden email]
> > > Web: www.stanford.edu/group/brunet/
> > > <Capture d’écran 2014-07-28 à 7.41.03 PM.png>------------------------------------------------------------------------------
> > > Infragistics Professional
> > > Build stunning WinForms apps today!
> > > Reboot your WinForms applications with our WinForms controls.
> > > Build a bridge from your legacy apps to the future.
> > > http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> > > Gmod-gbrowse mailing list
> > > [hidden email]
> > > https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
> >
> >
> >
> >
> > --
> > Bérénice A. BENAYOUN, Ph.D.
> > Stanford University/Genetics Department
> > BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> > M312 Alway Building
> > 300, Pasteur Drive
> > MC 5120
> > Stanford, CA 94305-5120
> > USA
> > Email: [hidden email]
> > Web: www.stanford.edu/group/brunet/
> > ------------------------------------------------------------------------------
> > Infragistics Professional
> > Build stunning WinForms apps today!
> > Reboot your WinForms applications with our WinForms controls.
> > Build a bridge from your legacy apps to the future.
> > http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> > Gmod-gbrowse mailing list
> > [hidden email]
> > https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
>
>
> ------------------------------------------------------------------------------
> Infragistics Professional
> Build stunning WinForms apps today!
> Reboot your WinForms applications with our WinForms controls.
> Build a bridge from your legacy apps to the future.
> http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
> _______________________________________________
> Gmod-gbrowse mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
>



--
------------------------------------------------------------------------
Scott Cain, Ph. D.                                   scott at scottcain dot net
GMOD Coordinator (http://gmod.org/)                     <a href="tel:216-392-3087" value="+12163923087" target="_blank">216-392-3087
Ontario Institute for Cancer Research



--
Bérénice A. BENAYOUN, Ph.D.
Stanford University/Genetics Department
BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
M312 Alway Building
300, Pasteur Drive
MC 5120
Stanford, CA 94305-5120

USA
Email: [hidden email]
Web: www.stanford.edu/group/brunet/



--
------------------------------------------------------------------------
Scott Cain, Ph. D.                                   scott at scottcain dot net
GMOD Coordinator (http://gmod.org/)                     216-392-3087
Ontario Institute for Cancer Research

------------------------------------------------------------------------------
Infragistics Professional
Build stunning WinForms apps today!
Reboot your WinForms applications with our WinForms controls.
Build a bridge from your legacy apps to the future.
http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
_______________________________________________
Gmod-gbrowse mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
Reply | Threaded
Open this post in threaded view
|

Re: lncRNA don't display correctly on GBrowse ?

Bérénice Benayoun
Hi Scott !

Thank you so much for your help ! I regenerated the gff3, making the lncRNA be the parent of the exons instead of the genes, and adding the subpart=exon. It finally worked, and looks great !

Again, thanks to everyone for super fast and useful feedback !!!

Best,

Berenice


2014-07-29 13:52 GMT-07:00 Scott Cain <[hidden email]>:
Hi Berenice,

WormBase is just using the standard gene glyph.  I think the problem is with your GFF: it has the parent of the exons being the gene, whereas the parent should be the mRNA (which is also a bit of a problem--clearly these aren't mRNAs either but that is a minor detail).  If you had explicit introns in the gff, I would suggest adding the "ignore_sub_part = intron" option, but it doesn't look like that's the case.

Scott



On Tue, Jul 29, 2014 at 4:41 PM, Bérénice Benayoun <[hidden email]> wrote:
Hi Scott and Michael,

Unfortunately, writing a derived class is beyond me (I look at the existing ones and I am unable to understand what they do without more comments). I tried doing just the UTR/exon substitution to make a lincRNA class, but I only got a red band with an error message in the browser.

The problem is that I also want to avoid putting CDSs where it doesn't make biological sense (there are obviously non coding...).

I had seen such tracks in UCSC and Wormbase (such as sent by Scott), so I thought it should be such a problem to get them to work in a new GBrowse derived browser.

Here is an extract of my non-coding GFF:

GapFilledScaffold_1264 Cufflinks gene 64907 65350 1000 - . ID=979895;Name=NFURLNR01264041000;Note=Cufflinks predicted non-coding
GapFilledScaffold_1264 Cufflinks mRNA 64907 65350 1000 - . ID=979896;Name=NFURLNR01264041000-mRNA-1;Parent=979895;Note=Cufflinks predicted non-coding
GapFilledScaffold_1264 Cufflinks exon 64907 64931 1000 - . ID=979897;Name=NFURLNR01264041000-mRNA-1:exon:1;Parent=979895;Note=Cufflinks predicted non-coding
GapFilledScaffold_1264 Cufflinks exon 65221 65350 1000 - . ID=979898;Name=NFURLNR01264041000-mRNA-1:exon:2;Parent=979895;Note=Cufflinks predicted non-coding


It has the parents relationships that should explain inheritance/parts to the glyph?

Maybe Wormbase has derived their own glyphs for non-coding ? Otherwise maybe my configurations are wrong ? Do you know if I can put my hands on the track configuration they use to see how it differs from mine ?

I looked through the code of processed_transcript and it seems to also require CDSs. With trasncript2, the best I can get is just vertical separation (see attached screencap) but the introns are displayed same as exons.

Sorry for my beginner questions - I am just really at a loss...

Thank you so much for your help !!!

Berenice



2014-07-29 12:03 GMT-07:00 Scott Cain <[hidden email]>:

Hi Berenice,

It should really work.  The sample of GFF below is from WormBase (I trimmed it a little to make it easier to read).

III  .  gene    832984  833347  .   +   .   ID=Gene:WBGene00219709
III  .   lincRNA 832984  833347  .   +   .       ID=Transcript:K02F3.14;Parent=Gene:WBGene00219709
III   .  exon    832984  833061  .   +   .   Parent=Transcript:K02F3.14
III   .  exon    833124  833347  .   +   .   Parent=Transcript:K02F3.14

I assume yours looks similar, right?  This lincRNA gene is represented in GBrowse in both the Curated Genes and Curated Genes (noncoding) tracks:


The coding genes in the Curated Genes track are gene->mRNA->CDS/UTR/exon (but it's clearly using the CDS/UTR data to draw the glyphs since it's coloring the UTRs appropriately).  There's nothing special about the track config, though it does have box_subparts turned on--I'm not sure what it's doing in this context.

Scott





On Tue, Jul 29, 2014 at 1:50 PM, Michael Dondrup <[hidden email]> wrote:
>
> Hi,
>
> I think this is due to the way the Bio::Graphics::Glyph::gene glyph works, you can have a look at its perldoc at http://search.cpan.org/~lds/Bio-Graphics/lib/Bio/Graphics/Glyph/gene.pm.
> The glyph works correctly only in case the Sequence ontology terms follow the exact hierarchy "gene"->"mRNA"->"exon"->"three_prime_utr"/ "five_prime_utr".
> The glyph doesn't resolve the SO hierarchy to find if e.g. rna_gene or lncrna_gene is also derived from gene. If it doesn't understand the type, it defaults to the
> 'standard' glyph which is just a block. There are two possibilities I can think of:
>
> - edit your GFF file, replace all SO terms by the ones that Bio::Graphics::Glyph::gene uses and update your features
> - derive your own glyph class from the source code of Bio::Graphics::Glyph::gene e.g. Bio::Graphics::Glyph::lncrna_gene and
> replace all occurrences  of SO terms in the code, most or all are found in regular expressions.
>
> Hope this helps
> Michael
>
> Michael Dondrup
> Postdoctoral fellow
> Sea Lice Research Centre/Department of Informatics
> University of Bergen
> Thormøhlensgate 55, N-5008 Bergen,
> Norway
>
> On Jul 29, 2014, at 7:20 PM, Bérénice Benayoun wrote:
>
> > Hi Todd,
> >
> > thanks for the quick answer !
> >
> > Actually, this was a previous try when I was trying to get this to work, and I commented it out to see if it made any difference... I de-commented it again, but no change, the multi-exon lncRNA is still one block...
> >
> > is there any hack or any other glyph that I could use that could at least show the exons connected but distinct ? I tried segments, but it also gave me a big block (though there were some vertical lines in there)...
> >
> > Thank you so much for your help, I really appreciate it !
> >
> > best,
> >
> > Berenice
> >
> >
> > 2014-07-28 21:19 GMT-07:00 Todd Harris <[hidden email]>:
> > Hi Berenice -
> >
> > What happens if you uncomment the “sub_part” option? That should force these entries to use exons as the subpart instead of CDS. Here’s an excerpt from the gene glyph documentation:
> >
> > This glyph is designed to work properly with GFF3-style three-tier
> > genes, in which the top level feature has the Sequence Ontology type
> > of "gene", the second level feature(s) have the SO type "mRNA", and
> > the third level feature(s) have the SO type "CDS", "five_prime_utr"
> > and "three_prime_utr."  Subparts named "UTR" are also honored.  The
> > feature can contain other subparts as well (e.g. exon, intron,
> > translation), but they are currently ignored unless the option
> > sub_part is supplied.  If the sub_part option is used that feature
> > type will be used and CDS and UTR features will be excluded.
> > This could be used for specifying that exons be used instead,
> > for example.
> >
> > Todd
> >
> >
> >
> > On Jul 29, 2014, at 9:43 AM, Bérénice Benayoun <[hidden email]> wrote:
> >
> > > Dear GBrowse developers,
> > >
> > > I have been trying to display predicted lncRNA of my favorite species (a fish whose genome I am annotating). I am running Gbrowse out of an AWS machine using the public image as a starting point.
> > >
> > > Everything is working well, except for lncRNAs: I can't get them to properly display the exon/intron structure (and the gene glyph doesn't look the same as for coding genes, through I configured everything identically save for the color). I attached a screen shot example...
> > >
> > > I created a gff3 file derived from my Cufflinks results (I generated the gene, mRNA and exon features, similarly to my coding genes, but no CDS, since these are non-coding). I have loaded them using a SQLite backend.
> > >
> > > Here is my invocation of the database in my conf file:
> > >
> > > [lncRNA:database]
> > > db_adaptor    = Bio::DB::SeqFeature::Store
> > > db_args       = -adaptor DBI::SQLite
> > >                 -dsn     /opt/gbrowse/databases/NotFur/Nfur_lnc.sqlite
> > > search options = default
> > >
> > > Here is the track invocation:
> > > [GenesNC]
> > > database           = lncRNA
> > > feature            = gene
> > > glyph              = gene
> > > #sub_part           = exon
> > > height             = 15
> > > bgcolor            = skyblue
> > > connector_color    = midnightblue
> > > decorate_introns   = 1
> > > description        = 1
> > > label              = 1
> > > category           = Genes
> > > key                = lncRNA genes
> > > link               = AUTO
> > >
> > > What am I doing wrong ? Is there any way to display non coding RNAs correctly ?
> > >
> > > Thank you so much in advance for your help - I've spent hours on it without progress and this is driving me insane !!!
> > >
> > > Best,
> > >
> > > Berenice
> > >
> > >
> > > --
> > > Bérénice A. BENAYOUN, Ph.D.
> > > Stanford University/Genetics Department
> > > BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> > > M312 Alway Building
> > > 300, Pasteur Drive
> > > MC 5120
> > > Stanford, CA 94305-5120
> > > USA
> > > Email: [hidden email]
> > > Web: www.stanford.edu/group/brunet/
> > > <Capture d’écran 2014-07-28 à 7.41.03 PM.png>------------------------------------------------------------------------------
> > > Infragistics Professional
> > > Build stunning WinForms apps today!
> > > Reboot your WinForms applications with our WinForms controls.
> > > Build a bridge from your legacy apps to the future.
> > > http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> > > Gmod-gbrowse mailing list
> > > [hidden email]
> > > https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
> >
> >
> >
> >
> > --
> > Bérénice A. BENAYOUN, Ph.D.
> > Stanford University/Genetics Department
> > BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
> > M312 Alway Building
> > 300, Pasteur Drive
> > MC 5120
> > Stanford, CA 94305-5120
> > USA
> > Email: [hidden email]
> > Web: www.stanford.edu/group/brunet/
> > ------------------------------------------------------------------------------
> > Infragistics Professional
> > Build stunning WinForms apps today!
> > Reboot your WinForms applications with our WinForms controls.
> > Build a bridge from your legacy apps to the future.
> > http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk_______________________________________________
> > Gmod-gbrowse mailing list
> > [hidden email]
> > https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
>
>
> ------------------------------------------------------------------------------
> Infragistics Professional
> Build stunning WinForms apps today!
> Reboot your WinForms applications with our WinForms controls.
> Build a bridge from your legacy apps to the future.
> http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
> _______________________________________________
> Gmod-gbrowse mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
>



--
------------------------------------------------------------------------
Scott Cain, Ph. D.                                   scott at scottcain dot net
GMOD Coordinator (http://gmod.org/)                     <a href="tel:216-392-3087" value="+12163923087" target="_blank">216-392-3087
Ontario Institute for Cancer Research



--
Bérénice A. BENAYOUN, Ph.D.
Stanford University/Genetics Department
BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
M312 Alway Building
300, Pasteur Drive
MC 5120
Stanford, CA 94305-5120

USA
Email: [hidden email]
Web: www.stanford.edu/group/brunet/



--
------------------------------------------------------------------------
Scott Cain, Ph. D.                                   scott at scottcain dot net
GMOD Coordinator (http://gmod.org/)                     <a href="tel:216-392-3087" value="+12163923087" target="_blank">216-392-3087
Ontario Institute for Cancer Research



--
Bérénice A. BENAYOUN, Ph.D.
Stanford University/Genetics Department
BRUNET Laboratory, 'Molecular Basis of Longevity and Age Related Diseases'
M312 Alway Building
300, Pasteur Drive
MC 5120
Stanford, CA 94305-5120

USA
Email: [hidden email]
Web: www.stanford.edu/group/brunet/

------------------------------------------------------------------------------
Infragistics Professional
Build stunning WinForms apps today!
Reboot your WinForms applications with our WinForms controls.
Build a bridge from your legacy apps to the future.
http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
_______________________________________________
Gmod-gbrowse mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse