FusionMap Change Log

From Array Suite Wiki

Jump to: navigation, search


Check FusionMap page for installation/example.

Contents

03/31/2015

Optimized for samples with >100 million reads; previously, it is slow

01/01/2015

Happy new year! New year 2015 release.

10/17/2014

Change to use OSA version=4

01/04/2014

Happy new year! Wish you have a wonderful year 2014.

If your FusionMap reports "this version of software has expired", we have updated FusionMap with latest configuration. Please Download it again.

If you have not updated your FusionMap for a while, please read previous update notes.

In the new version, we have add additional fusion report columns:

07/30/2013

If your old mono is not working with FusionMap, we recommend using MONO Version 2.10.9 or higher for latest FusionMap release.

Add more fusion report annotations, such as SplicePatternClass (Canonical or NonCanonical), FrameShiftClass (Frame shift or inFrame), Distance (Gap size between two breakpoint or -1 if located on two different chromosomes) and OnExonBoundary (Both, Single or None), for users to further filter false positives.

Note: By default, jobs will be spawn off and multiple samples will be run in parallel in individual process. However, for some machines, the path to mono are not remembered and they cannot start. To solve the problem, please add MonoPath in <Options> section, such as

<Options>
MonoPath=/opt/bin/mono
PairedEnd=True
RnaMode=True

12/08/2012

bug fix and software update

Support BAM file input for semap. FusionMap will use unmapped reads in the BAM files to do fusion alignment and detection. See example of running FusionMap semap on BAM file

10/02/2012

bug fix and software update

08/16/2012

bug fix

04/16/2012

Fix a bug dealing with the situation when read1 and read2 have the same name without /1 /2 or :1 :2.

Adding instruction to use OSA + FusionMap::pereport.

03/03/2012

New release FusionMap.2012-03-03

  • Use OSA as the basic aligner for FusionMap;
  • Resolve some Memory problem using MMF;
  • Resolve some I/O problem related to default Linux flushing mechanism

If your old mono is not working with FusionMap, we recommend using MONO Version 2.10.8 or higher for latest FusionMap release.

Direct link to MONO 2.10.8: http://download.mono-project.com/sources/mono/mono-2.10.8.tar.gz Use mono instead of mono-sgen in the latest MONO+FusionMap

01/01/2012

Happy new year! Wish you have a wonderful year 2012.

If your FusionMap reports "this version of software has expired", we have updated FusionMap with latest configuration. Please Download it again.

If you have not updated your FusionMap for a while, please read previous update notes.

11/22/2011

Allow users to specify the temp folder using a line (TempPath=/tmp in the <output> section) in the config/control file. We notice that some users are trying to run the FusionMap on a cluster and had file handler problem in tmp file I/O. User can solve this problem by specifying a Temp folder. NOTE Everything in the TempPath will be deleted in the end of FusionMap. Please create a folder (or check whether a folder exist), e.g. /tmp/FusionMapTmp, before you run FusionMap in the pipeline. It could be dangerous to wipe the whole /tmp folder in linux.

Adding Use32BitMode=False to enforce 64bit running mode.

Predict the 5' and 3' fusion gene (A->B, BCR->ABL1), output the splice pattern and open reading frame status

10/09/2011

v2.1 has been released.

performance improvement update (no bug fix)

Find a bug in MONO 2.10.x. Please use FusionMap under MONO 2.8.1 now.

Direct link to MONO 2.8.1: http://ftp.novell.com/pub/mono/sources/mono/mono-2.8.1.tar.bz2

08/12/2011

v2.0 has been released with the following changes:

  • 2x faster than before (especially for linux and thread>4 performance)
  • Automatically clean temporary files when exits
  • Output SAM file for paired end fusion report

RECOMENDATION: Apply FusionMap directly to the whole dataset In the paper, we described a pre-processing step filtering with a regular read mapper, using Bowtie or Tophat. After that, these unmapped reads are used to run FusionMap. However, the preprocessing before applying FusionMap is optional. We later find that it is too complicated to ask user to find unmapped reads. We are now recommending user to run FusionMap directly on the whole dataset since it won't take too much time now. For paired-end reads, we added PairedEnd option in FusionMap, if user specify PairedEnd=True, FusionMap will pair two files and consider them as NGS raw reads from the same sample.

05/20/2011

v 1.42.1 has been released with the following changes: add PairedEnd option.

FusionMap is not limited to a single dataset, but rather can be applied to multiple NGS datasets in a single run. It provides an easy way to detecting recurrent fusion events by grouping fusion junctions by rows in one table. By default, PairedEnd=False, FusionMap considers each NGS file as one sample. If user input two NGS files, FusionMap runs fusion detection twice in a single run and generates one report grouping fusion junctions by rows in one table.

If user specify PairedEnd=True, FusionMap will pair two files and consider them as NGS reads from the same sample. It provides convenience if user wants to detection fusion junctions using the raw paired-end datasets.

05/11/2011

FusionMap paper has been accepted by Bioinformatics. Please cite as below:

FusionMap: detecting fusion genes from next-generation sequencing data at base-pair resolution

Huanying Ge; Kejun Liu; Todd Juan; Fang Fang; Matthew Newman; Wolfgang Hoeck

Bioinformatics (2011) 27 (14): 1922-1928. doi: 10.1093/bioinformatics/btr310 BibTex format

04/15/2011

v 1.42 has been released with the following changes:

  1. Clean the parameters in control file. User only needs to specify the parameters that are different from the default values
  2. Applying a simple check rule on unmapped reads for seed fusion read selection. In the previous version, FusionMap cut and aligned two ends of each unmapped read to the reference. In this version, FusionMap aligns the first and last α-nt (by default α=25 for 75nt reads) to the reference. If both α-nt short ends are mapped, FusionMap will choose it as a candidate seed read and then cut the read at all positions in the seed region to find an optimal cutting position for the fusion junction detection. This simple check speeds up the running time, particularly for large NGS dataset (>40 million reads).
  3. Output the alignment of fusion reads in SAM format which can be visualized in genome browsers (e.g. IGV) (see (4) Output files).

02/23/2011

v 1.30 has been released: added advanced filtering based on gene black list and gene family list; further increased the specificity.

11/30/2010

The original website for FusionMap has been setup. Version 1.0 has been released. A getting started instruction webpage is included for your convenience.