MSIsensor is a C++ program to detect replication slippage variants at microsatellite regions, and differentiate them as somatic or germline. Given paired tumor and normal sequence data, it builds a distribution for expected (normal) and observed (tumor) lengths of repeated sequence per microsatellite, and compares them using Pearson's Chi-Squared Test. Comprehensive testing indicates MSIsensor is an efficient and effective tool for deriving MSI status from standard tumor-normal paired sequence data. Since there are many users complained that they don't have paired normal sequence data or related normal sequence data can be used to build a paired normal control, we released MSIsensor2, and which is specially designed for MSI detection using tumor only or ctDNA sequencing data. Our test results show that the performance of MSIsensor2 is comparable with paired tumor and normal sequence data input. In particular, the MSIsensor2 is 10 times faster than the MSIsensor. A typical WES data can be finished within 180 seconds (test on both hg19 and hg38 bams). Please try the MSIsensor2 here: https://github.com/niu-lab/msisensor2 or require any further details here: http:https://niulab.scgrid.cn/msisensor2/index.html .

If you used this tool for your work, please cite PMID 24371154

Beifang Niu*, Kai Ye*, Qunyuan Zhang, Charles Lu, Mingchao Xie, Michael D. McLellan, Michael C. Wendl and Li Ding#.MSIsensor: microsatellite instability detection using paired tu-mor-normal sequence data. Bioinformatics 30, 1015–1016 (2014).

You may already have these prerequisite packages. If not, and you're on Debian or Ubuntu:

sudo apt-get install zlib1g-dev libncurses5-dev libncursesw5-dev

If you are using Fedora, CentOS or RHEL, you'll need these packages instead:

sudo yum install zlib-devel ncurses-devel ncurses

• For Linux and OSX binaries, look for msisensor.linux and/or msisensor.macos in attachments to each release

conda install msisensor

Clone the msisensor master branch, and build the msisensor binary:

git clone https://github.com/ding-lab/msisensor.git
cd msisensor
make

Now you can put the resulting binary where your $PATH can find it. If you have su permissions, then we recommend dumping it in the system directory for locally compiled packages:

sudo mv msisensor /usr/local/bin/

    Version 0.6
    Usage:  msisensor <command> [options]

Key commands:

    scan            scan homopolymers and miscrosatelites
    msi             msi scoring

msisensor scan [options]:

   -d   <string>   reference genome sequences file, *.fasta format
   -o   <string>   output homopolymer and microsatelittes file

   -l   <int>      minimal homopolymer size, default=5
   -c   <int>      context length, default=5
   -m   <int>      maximal homopolymer size, default=50
   -s   <int>      maximal length of microsate, default=5
   -r   <int>      minimal repeat times of microsate, default=3
   -p   <int>      output homopolymer only, 0: no; 1: yes, default=0

   -h   help

msisensor msi [options]:

   -d   <string>   homopolymer and microsates file
   -n   <string>   normal bam file
   -t   <string>   tumor  bam file
   -o   <string>   output distribution file

   -e   <string>   bed file, optional
   -f   <double>   FDR threshold for somatic sites detection, default=0.05
   -c   <int>      coverage threshold for msi analysis, WXS: 20; WGS: 15, default=20
   -z   <int>      coverage normalization for paired tumor and normal data, 0: no; 1: yes, default=0
   -r   <string>   choose one region, format: 1:10000000-20000000
   -l   <int>      minimal homopolymer size, default=5
   -p   <int>      minimal homopolymer size for distribution analysis, default=10
   -m   <int>      maximal homopolymer size for distribution analysis, default=50
   -q   <int>      minimal microsates size, default=3
   -s   <int>      minimal microsates size for distribution analysis, default=5
   -w   <int>      maximal microstaes size for distribution analysis, default=40
   -u   <int>      span size around window for extracting reads, default=500
   -b   <int>      threads number for parallel computing, default=1
   -x   <int>      output homopolymer only, 0: no; 1: yes, default=0
   -y   <int>      output microsatellite only, 0: no; 1: yes, default=0

   -h   help

1. Scan microsatellites from reference genome:

    msisensor scan -d reference.fa -o microsatellites.list
   

2. MSI scoring:

    msisensor msi -d microsatellites.list -n normal.bam -t tumor.bam -e bed.file -o output.prefix
   

   Note: normal and tumor bam index files are needed in the same directory as bam files

The list of microsatellites is output in "scan" step. The MSI scoring step produces 4 files:

    output.prefix
    output.prefix_dis_tab
    output.prefix_germline
    output.prefix_somatic

1. microsatellites.list: microsatellite list output ( columns with *_binary means: binary conversion of DNA bases based on A=00, C=01, G=10, and T=11 )

    chromosome      location        repeat_unit_length     repeat_unit_binary    repeat_times    left_flank_binary     right_flank_binary      repeat_unit_bases      left_flank_bases       right_flank_bases
    1       10485   4       149     3       150     685     GCCC    AGCCG   GGGTC
    1       10629   2       9       3       258     409     GC      CAAAG   CGCGC
    1       10652   2       2       3       665     614     AG      GGCGC   GCGCG
    1       10658   2       9       3       546     409     GC      GAGAG   CGCGC
    1       10681   2       2       3       665     614     AG      GGCGC   GCGCG
   

2. output.prefix: msi score output

    Total_Number_of_Sites   Number_of_Somatic_Sites %
    640     75      11.72
   

3. output.prefix_dis_tab: read count distribution (N: normal; T: tumor)

    1       16248728        ACCTC   11      T       AAAGG   N       0       0       0       0       1       38      0       0       0       0       0       0       0
    1       16248728        ACCTC   11      T       AAAGG   T       0       0       0       0       17      22      1       0       0       0       0       0       0
   

4. output.prefix_somatic: somatic sites detected ( FDR: false discovery rate )

    chromosome   location        left_flank     repeat_times    repeat_unit_bases    right_flank      difference      P_value    FDR     rank
    1       16200729        TAAGA   10      T       CTTGT   0.55652 2.8973e-15      1.8542e-12      1
    1       75614380        TTTAC   14      T       AAGGT   0.82764 5.1515e-15      1.6485e-12      2
    1       70654981        CCAGG   21      A       GATGA   0.80556 1e-14   2.1333e-12      3
    1       65138787        GTTTG   13      A       CAGCT   0.8653  1e-14   1.6e-12 4
    1       35885046        TTCTC   11      T       CCCCT   0.84682 1e-14   1.28e-12        5
    1       75172756        GTGGT   14      A       GAAAA   0.57471 1e-14   1.0667e-12      6
    1       76257074        TGGAA   14      T       GAGTC   0.66023 1e-14   9.1429e-13      7
    1       33087567        TAGAG   16      A       GGAAA   0.53141 1e-14   8e-13   8
    1       41456808        CTAAC   14      T       CTTTT   0.76286 1e-14   7.1111e-13      9
   

5. output.prefix_germline: germline sites detected

    chromosome   location        left_flank     repeat_times    repeat_unit_bases    right_flank      genotype
    1       1192105 AATAC   11      A       TTAGC   5|5
    1       1330899 CTGCC   5       AG      CACAG   5|5
    1       1598690 AATAC   12      A       TTAGC   5|5
    1       1605407 AAAAG   14      A       GAAAA   1|1
    1       2118724 TTTTC   11      T       CTTTT   1|1
   

We provided one small dataset (tumor and matched normal bam files) to test the msi scoring step:

    cd ./test
    bash run.sh

We also provided a R script to visualize MSI score distribution of MSIsensor output. ( msi score list only or msi score list accompanied with known msi status). For msi score list only as input:

    R CMD BATCH "--args msi_score_only_list msi_score_only_distribution.pdf" plot.r

For msi score list accompanied with known msi status as input:

    R CMD BATCH "--args msi_score_and_status_list msi_score_and_status_distribution.pdf" plot.r

If you have any questions, please contact one or more of the following folks: Beifang Niu [email protected] Kai Ye [email protected] Li Ding [email protected]