Anyone wiht experience using MATS for alternative splicing?

LP_SEP23 · 09-14-2012, 10:24 AM

Hi,
I am looking into analyzing differential alternative splicing after a treatment. I have 2 conditions (control and treatment) and 4 biological replicates on each. I have on average 75 million pairs of 100bp per sample.

I am trying to use MATS do do this. I seem to have MATS working as per the test run. Although I could use my fastq files, I already have my reads mapped and wanted to use bam files, which were not generated by tophat and it is not working. Anyone knows what can be causing the problem? what does MATS expect from the bam fields?

Also, since I have to rent nodes form a cluster, what is the RAM required for MATS to do the analysis in mammalian size genomes with this number of replicates (my fastq files are ~17GB each, my bam files are ~9GB). For how long should I expect them to run?

thanks
Lucia

statsteam · 04-29-2014, 04:06 PM

As long as I know, MATS works with standard bam files with NH:i:x attribute.
For me, all MATS run took about 24 hours. Hope this helps.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
RNAseq Alternative Splicing	awk	Bioinformatics	33	06-10-2015 03:26 AM
alternative splicing of RNA-Seq	AronaldJ	RNA Sequencing	2	09-10-2012 11:35 AM
Alternative splicing with RNA Seq	slny	Bioinformatics	1	06-02-2011 05:44 AM
RPKM for alternative splicing genes?	baohua100	Bioinformatics	2	09-15-2009 12:57 AM

09-14-2012, 10:24 AM	#1
LP_SEP23 Junior Member Location: Philadelphia Join Date: Feb 2012 Posts: 9	Anyone wiht experience using MATS for alternative splicing? Hi, I am looking into analyzing differential alternative splicing after a treatment. I have 2 conditions (control and treatment) and 4 biological replicates on each. I have on average 75 million pairs of 100bp per sample. I am trying to use MATS do do this. I seem to have MATS working as per the test run. Although I could use my fastq files, I already have my reads mapped and wanted to use bam files, which were not generated by tophat and it is not working. Anyone knows what can be causing the problem? what does MATS expect from the bam fields? Also, since I have to rent nodes form a cluster, what is the RAM required for MATS to do the analysis in mammalian size genomes with this number of replicates (my fastq files are ~17GB each, my bam files are ~9GB). For how long should I expect them to run? thanks Lucia

04-29-2014, 04:06 PM	#2
statsteam Member Location: Californica Join Date: Sep 2009 Posts: 19	As long as I know, MATS works with standard bam files with NH:i:x attribute. For me, all MATS run took about 24 hours. Hope this helps.