JBROWSE project. Unsure if i should post to group.

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

JBROWSE project. Unsure if i should post to group.

Foley, Brian T
Dear Robert Buels et al,

      We (The Human Immunodeficiency Virus Databases at LANL) have been developing a "HIV Genome sites of Interest" genome map.
In prior years we had various methods of showing this type of information in text, images and spreadsheets,
and we moved to using JBROWSE This year.



Older files:




etc...


We are very happy with JBROWSE so far, and I am wondering about contributing to the 
[hidden email] list to share what we have done, and perhaps ideas for further work.
I participate in many on-line user groups, blogs, discussion forums and so on, and I know that each
one has it's own level of discourse which is considered acceptable or welcome.  I have been subscribed to the
jbrowse-dev list for several months, but I must admit that I do not read most posts there, so I am still
pretty ignorant of the "culture" on this list.  So, please let me know if I should be posting this somewhere
else.

HIV has probably the smallest genome where a "genome browser" would be useful.  It is roughly
9.7 kb long and contains 9 genes.

What is rather unique about HIV, is how deeply it has been studied from a number of different angles.
In addition to rich knowledge of each gene and protein in the genome, there is rich knowledge of how the
human immune system has interacted with the virus and driven the evolution of the virus over time.  Thus, in addition
to "tracks" for the genome, mapping where the genes are located, we have tracks showing which sites
are common targets of various immune responses (CD4+ CTL epitopes, CD8+ epitopes, Antibody epitopes), and
tracks for "rate of evolution" or "entropy".

    We are using JBROWSE, rather than Artemis of other Genome Browsers, largely because a large segment
of our user base is located in third world countries, and composed of HIV researchers who are not computer
scientists or trained in "bioinformatics" .  We want a tool that works within a web browser, such as Safari or
Firefox, and can then "link out" from any site of interest in the genome (such as a particular CTL epitope)
to our database records for that site.


for an example:

should show the Vif protein, with a CTL epitope:
KARGWFYRHHYESP

and right-clicking on that epitope pulls up a menu of choices taking the user to an alignment of the diversity found
in this epitope among global strains of the virus:

or to the IHIV Immunology Database record for the epitope:

     I am quite certain that the focus of your JBROWSE development efforts are for much larger genomes with
less dense annotation.  But I thought you all might be interested to see how we are using it.  

Brian T. Foley, PhD
HIV Databases
Los Alamos National Laboratory
[hidden email]
505 665-1970

Reply | Threaded
Open this post in threaded view
|

Re: JBROWSE project. Unsure if i should post to group.

Diesh, Colin M.
Hi Brian,
This looks great, thanks for sharing! We have a demos page at jbrowse.org that definitely needs some updates, send me an email if you want to share it on the website.

Also, I have to mention that Rob is not actively developing JBrowse anymore, but we still have an active group on GitHub. We welcome feature requests and bug reports, and even any code changes that you have made!

-Colin


On Mar 17, 2014, at 7:03 PM, Foley, Brian T <[hidden email]> wrote:

Dear Robert Buels et al,

      We (The Human Immunodeficiency Virus Databases at LANL) have been developing a "HIV Genome sites of Interest" genome map.
In prior years we had various methods of showing this type of information in text, images and spreadsheets,
and we moved to using JBROWSE This year.



Older files:




etc...


We are very happy with JBROWSE so far, and I am wondering about contributing to the 
[hidden email] list to share what we have done, and perhaps ideas for further work.
I participate in many on-line user groups, blogs, discussion forums and so on, and I know that each
one has it's own level of discourse which is considered acceptable or welcome.  I have been subscribed to the
jbrowse-dev list for several months, but I must admit that I do not read most posts there, so I am still
pretty ignorant of the "culture" on this list.  So, please let me know if I should be posting this somewhere
else.

HIV has probably the smallest genome where a "genome browser" would be useful.  It is roughly
9.7 kb long and contains 9 genes.

What is rather unique about HIV, is how deeply it has been studied from a number of different angles.
In addition to rich knowledge of each gene and protein in the genome, there is rich knowledge of how the
human immune system has interacted with the virus and driven the evolution of the virus over time.  Thus, in addition
to "tracks" for the genome, mapping where the genes are located, we have tracks showing which sites
are common targets of various immune responses (CD4+ CTL epitopes, CD8+ epitopes, Antibody epitopes), and
tracks for "rate of evolution" or "entropy".

    We are using JBROWSE, rather than Artemis of other Genome Browsers, largely because a large segment
of our user base is located in third world countries, and composed of HIV researchers who are not computer
scientists or trained in "bioinformatics" .  We want a tool that works within a web browser, such as Safari or
Firefox, and can then "link out" from any site of interest in the genome (such as a particular CTL epitope)
to our database records for that site.


for an example:

should show the Vif protein, with a CTL epitope:
KARGWFYRHHYESP

and right-clicking on that epitope pulls up a menu of choices taking the user to an alignment of the diversity found
in this epitope among global strains of the virus:

or to the IHIV Immunology Database record for the epitope:

     I am quite certain that the focus of your JBROWSE development efforts are for much larger genomes with
less dense annotation.  But I thought you all might be interested to see how we are using it.  

Brian T. Foley, PhD
HIV Databases
Los Alamos National Laboratory
[hidden email]
505 665-1970


Reply | Threaded
Open this post in threaded view
|

JBROWSE project to share on "demos" page. And request for a "search" function.

Foley, Brian T
The HIV Databases would be happy to serve as an example on the DEMOS page:

Our genome browser is a "work in progress", but I think we have some features already
working that could be useful for others to see.  

I am more familiar with Artemis than most other genome browsers, and with that I have
looked mostly at bacterial and viral genomes rather than eukaryotic chromosomes.

When I do "surf around" in eukaryote chromosomes, such as:


or 


I am a bit frustrated that I can't easily figure out how to zoom in on a particular gene of interest.
Even within GenBank/NCBI, I know how to find a gene of interest with BLAST or text searching.,
for example to find the eEF-2 gene on human chromosome 19 bases 3976056..3985463, complement

or 
   gene            complement(3974658..3985048)
                     /gene="EEF2"
but when I go to a site such as "Genomes Unzipped"   

and load up chromosome 19 and zoom in to that region of the chromosome, 
I can't find this gene, perhaps because they are using a different chromosome 19 assembly?
The Genomes Unzipped browser has a RefSeq Genes track, but I don't see any way to search
in that track for eEF-2  or Elongation.

It would be very useful to have a way to search the sequence for my sequence (DNA or protein) of interest, or search
the annotated genes track for the name of my gene.

Sincerely,
Brian Foley
HIV Databases


On Mar 18, 2014, at 10:32 AM, Diesh, Colin M. wrote:

Hi Brian,
This looks great, thanks for sharing! We have a demos page at jbrowse.org that definitely needs some updates, send me an email if you want to share it on the website.

Also, I have to mention that Rob is not actively developing JBrowse anymore, but we still have an active group on GitHub. We welcome feature requests and bug reports, and even any code changes that you have made!

-Colin


On Mar 17, 2014, at 7:03 PM, Foley, Brian T <[hidden email]> wrote:

Dear Robert Buels et al,

      We (The Human Immunodeficiency Virus Databases at LANL) have been developing a "HIV Genome sites of Interest" genome map.
In prior years we had various methods of showing this type of information in text, images and spreadsheets,
and we moved to using JBROWSE This year.



Older files:




etc...


We are very happy with JBROWSE so far, and I am wondering about contributing to the 
[hidden email] list to share what we have done, and perhaps ideas for further work.
...

Brian T. Foley, PhD
HIV Databases
Los Alamos National Laboratory
[hidden email]
505 665-1970

Reply | Threaded
Open this post in threaded view
|

Re: JBROWSE project to share on "demos" page. And request for a "search" function.

Foley, Brian T
How embarrassing.    

I was able to find the gene in the Genomes Unzipped browser, by searching on eEF2 in the "search box".
I find the gene at 3,925,180 .. 3,938,349

I had only tried eEF-2 and elongation before.

But still it might be nice to be able to search with an amino acid
string, a DNA string, or text words in a specific track.

Brian


On Mar 18, 2014, at 11:52 AM, Foley, Brian T wrote:

The HIV Databases would be happy to serve as an example on the DEMOS page:

...

but when I go to a site such as "Genomes Unzipped"   
... I don't see any way to search
in that track for eEF-2  or Elongation.


Brian T. Foley, PhD
HIV Databases
Los Alamos National Laboratory
[hidden email]
505 665-1970