Hello from Boston (IGV creator)

[Jim Robinson]

Hi, I'm the creator if IGV (Integrative Genomics Viewer) and am joining the forum to keep abreast of active issues, takes suggestions from this community, and hopefully get some of my questions answered.

IGV can be used to visualize short-read alignments, as a generic genome browser, and for microarray based experiments. I'm still learning my way around this website, and will try to subscribe to some of the forums, but fill free to send me a private message re IGV questions.

Jim Robinson
Broad Institute

[sdarko]

Hi Jim,

I caught your presentation at NIH on IGV recently. Great tool and I've been using it ever since.

Sam

[polivares]

I want to thank you for the great tool you have developed !

You make me save LOTS of time.

All the best

Pedro

[Buzz0r]

hey there,
i am thinking about using the IGV for my research but i am also considering the new SAVANT. is there a video of the talk you gave earlier this year at NIH? or any other presentation of the IGV?

thanks, thomas

[Jim Robinson]

Hi Thomas,

Sorry but we don't have any videotapes of presentations. I suggest you try both tools, choice is good and you might find you like both for different things.

-- Jim

[Buzz0r]

okay,

we already are looking at both and see nice visual possibilities in either one. We are only using the viewers for inspection due to our own analytic and statistic scripts so i cannot comment on the further possibilitier like the analytical plugins for example.

But for other readers: both are very easy to handle and very good on the memory usage. I will update you if i find any big runtime, memory aso differences if there are any readers who are actually interested in this IGV vs. SAVANT comparison. I might also look at another viewer called "SVA" if i ever get enough spare time soon...schedule is packed as usual ...

best, thomas

[Jim Robinson]

Very good. If you test any really large data files, such as "wig" files, you should use "igvtools" to convert the ascii file to a .tdf file. These are more or less equivalent in purpose to the ".savant" files, but it is optional to create these in IGV so many people don't know about them. You can run igvtools from the command line or launch a GUI to run them from the file menu. This format is used for all the large tracks in our "hg18" load from server, such as the Encode data.

[boetsie]

<<<MOVED THIS QUESTION TO Bioinformatics >>>

Hi Jim,

I'm currently testing IGV for visualising SNPs on a bacterial genome. I've generated a SNP table with CLC which looks like this;

Reference Position Consensus Position Variation Type Length Reference Variants Allele Variations Frequencies Counts
122 122 SNP 1 G 1 A 100 133
1411 1411 SNP 1 T 1 C 100 135
1639 1639 SNP 1 G 1 A 100 127
1738 1738 SNP 1 A 1 G 100 118
....

I have converted this file to a simple .bed file;

GENOMENAME 121 122 A
GENOMENAME 1410 1411 C
GENOMENAME 1638 1639 A
GENOMENAME 1737 1738 G
......

This all works, as i can read in the file and visualize it along the genome. However, i want to add more information, like the 'Frequences' and 'Counts' column. These columns should be displayed if i scroll over a SNP.

How can i do this? How should the input look like?

Thanks for the help,

Boetsie

[Jim Robinson]

Hi,

Unfortunately the bed format is pretty rigid in column definitions. One alternative is to us GFF3 (http://gmod.org/wiki/GFF3#GFF3_Format) and use the last column (attributes). So

1411 1411 SNP 1 T 1 C 100 135

would become

chrName . SNP 1411 1411 . . . ID=SNP1411;Name=C;Reference=T; etc

GFF is very verbose, you will have to repeat the attribute names on every row. You could also use VCF format, which IGV supports and is defined for this. Finally, I could add the snp table format, email a reference to the spec to [email protected] and we can discuss it further.

Regards,

Jim

[Jean]

Hi Jim,
I have been using IVG to view bacterial RNA-seq data and I really like the viewer. However, I am having trouble getting the gene annotation information to display. I have tried loading the GFF file with the genome FASTA but the track is blank - am I possibly doing something wrong?

Jean

[Jim Robinson]

Hi Jean,

The most common cause of this is a mismatch in sequence (chromosome/contig/whatever) names between the fasta and GFF file. If you want to email us a sample of your GFF and fasta file I'll look at it in more detail. Send it to [email protected].

-- Jim

[gaffa]

Hi Jim, I really like the IGV browser but one thing that could use some improvement in my opinion is the display of insertions. I don't know if there are any plans to change this, but the current system with the "I"-symbol marking the location of an insertion doesn't really allow for detailed examination of sequence context in/around insertions. Instead I usually resort to samtools tview for such examination.

[Jim Robinson]

Yes, that's embarrassing really. It was copied from UCSC's method of displaying insertions in multiple alignments until something better could be coded. Its high on my list to improve.

So the plan is to "stretch" the reference where insertions occur so that the whole sequence of the insertion can be seen. The reference sequence will just show dots there. It really is high on my list but there's a flood of other things being implemented at the moment, I expect to get to this within the next 6-8 weeks however.

-- Jim

[Jean]

Hi Jim.
I just wanted to thank you for the update to IGV to display my custom GFF file. We now have some really great visualization of our transcriptome mappings that will be invaluable for analysis.

Attached is a screenshot for anyone who might be interested.

I've also noticed I can enter annotation information into the file which is displayed in IGV by mouse-over - very useful!

Jean

[doron]

Gff3 problems

Hi Jim,

great work on the IGV, it is very helpful!
I have a small problem with viewing gff3 data.
I attached an example of a gene i'm trying to view. (gene.gff3)
The viewer only shows 2 of 3 UTRs, with no CDSes.

I found a way to improve the situation by inserting an exon record (gene_improved.gff3)
But, In this example, the viewer shows only 1 of the 2 5UTRs.

Is there a way to show all UTRs (in narrow blocks) and Cdses (in thick blocks)?

Thank you very much!
Doron Shem-Tov