NEW: Alview - Fast alignment viewer ....

[Richard Finney]

This is fast, portable .... and in the Public Domain. No registration. Source is available. GUI executables for MAC, Linux 64 bit, Windows 64bit are available.

Alview ( ALignment VIEWer ) is a fast, compiled to native architecture computer program for visualizing the alignment of sequencing data. The inputs are files of short read sequences aligned to a reference genome in the SAM/BAM format. Outputs are images representing these short reads aligned to the genome. Alview is written in portable C with optional Graphical User Interface (GUI) code written in C++. Alview can be compiled to run in three different ways:

1) as a webserver (CGI, apache, probably works with NGINX)
2) as a command line tool (just batch up a script and review those calls fast !!!)
or
3) as a native graphical user interface executable. (Explore that BAM at the speed of your curiosity)

It is compatible with Microsoft Windows, Linux and Apple OS X.

COMMERCIAL REDISTRIBUTION OF GUI WILL REQUIRE PURCHASE OF Qt LICENSE. But that's only if you're forking it and and trying to make some bucks. Otherwise, do what you want to.

Availability:
Web Demo: https://cgwb.nci.nih.gov/cgi-bin/alview
Source and Windows/Mac/Linux packages: https://github.com/NCIP/alview



THIS IS BRAND NEW, SO I"D APPRECIATE ANY FEEDBACK AND WILL FIX IT PRONTO.

[Richard Finney]

Alview is a NATIVE GUI for browsing bam files. For GUI version, it now uses native libraries: Win32, GTK Linux, Cocoa for Mac. (Qt has been removed).

Windows Installer Version for Alview Now Available, along with Mac DMG and linux tarball.

Source: https://github.com/NCIP/alview
Download Packages for Linux/Windows/Mac : http://45.56.125.191/
Web Demo : https://cgwb.nci.nih.gov/cgi-bin/alview

This is much improved. Now statically compiled with only common system run time libraries required. Typically there is NO system or obscure library updates required.

Power users will appreciate the ability to generate pictures for hundreds of thousands of SNP calls and view them in a slide show using the command line (with some deliberate hand work to tie all together in a custom script you create).

[apredeus]

Is it faster/more efficient than IGV? What are the most useful features you'd recommend? We are starting to work on some exome data, and a good efficient alignment browser would be pretty great.

[Richard Finney]

Alview is faster as it's in native compiled C (and C++ for windows and Objective C for mac). There are separate implementations for various operating systems; static linking and usage of old, standard system libraries makes it highly portable to newer operating systems.

It is implemented by someone who doesn't want to get slowed down waiting for screen redraws.

You will find that IGV has many more features; however ... and actually delivers on portability to systems that support the Java Virtual Machine.

Please note the batch processing capabilities. Batch invocation produces an image of the alignment. This should be enough of hint for anybody who hacks around in a favourite GUI capable programming language as alview can be invoked as an external program ( i.e. a "system() call" in C/Unix). With some custom script-fu you can view thousands and thousands of snp calls ... and even do your own quality assessment accounting. The resulting images can , of course, also be clickable from html or Excel.

Despite some "off the richter scale" p-values, viewing the results might clue you into some "chemistry bias" in your calls. Sometimes the eye can spot these.

I'd be happy to *try* and add a feature if I can, so let me know.

[apredeus]

Sounds pretty good, thank you, we'll definitely give it a shot. Are there manuals or some examples of batch use? That does sound very useful.

[Richard Finney]

The Alview github site has a very old school asciii user's manual.

Calling a command line version with no args will give the current invocation. It's pretty easy.

Scripting and cranking out 1000s of images and managing them, however, is a non-trivial user's task;.

The customization of a users workflow and reviewing the results is often a powerful and worthwhile quality control check.

[gringer]

Something I'd like to see in an offline alignment viewer is a stacked 3D plot, where the sample is on the depth axis. See attached for an example done in GBrowse, which I think is a pre-generated track that can't be customised.

We've got about 70 different samples that would be great to be able to display at once (or a subset of ~12), and I think the stacked plot would make that achievable. Just a coverage plot in log space would be great, but I'd also be interested to know what an individual read plot would look like (trimming off any reads above ~100X coverage).

[shuoguo]

I would appreciate if you show more examples. I am in need of such tool and will try your tool when I got a chance.

Features I really would like are:
- work with whole genome BAMs
- show mismatches/insertions/deletions/soft-clips to reference genome, show gene tracks (one or all isoforms), and amino acids (good to have).
- show coverages on top of alignments
- show different shade for base quality scores (e.g. grey background for low base quality scores and black for high)
- show different strand in different groups
- show read names on left panels (so we can look at pair-end reads)

Thanks,

Shuoguo

[Richard Finney]

Help page is here:

http://45.56.125.191/help.html

This image for this is reproduced here :

img src="http://i.imgur.com/FUQR3Bg.png"

Whole genomes are supported.
Genetrack does not show isoforms, just all exons.
Grayscale base quality is there for most instantiations.
A read coverage track is not there, though it's now in the "to do" list.
Read names are available only in the web version with "samview" options.
This is a good suggestion for an improvement. I had not implemented it earlier as I wanted a very spartan "make the snp call as fast as possible" tool. In practice, Alview is best used as a fast tool to review dozens to hundred of thousand snp calls. I usually don't care what the read name is
when punching through all the snp calls in a project, but I'll see what I can do to make this easy to get.

Thank your for the feedback/requests; I'm overdue for pushing out the latest update and should get to it this week.