Yizhe Song [email protected]

& Daniel Cui Zhou [email protected]

Last updated: 04/11/22

Adapted from Daniel Cui Zhou's hg38 pipeline.

VERSION: v1.5 Build: hg38 Cluster: compute1

MSIsensor can be found here: https://github.com/ding-lab/msisensor. MSI scores can roughly be interpreted as the percentage of microsatellite sites (with deep enough sequencing coverage) that have a lesion.

Samples with an MSIscore > 3.5 are classified as "MSI-High" and the rest will be classified as "MSS." An intermediate class with 1.0 <= score <= 3.5 can be defined as "MSI-Low."

The site file annotation is as follows: 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 More details can be found in the github page.

Processing details: Run "makeDir_msi_v1.py" to generate a "to_run.sh" file and create the appropriate folder directory for each sample. Run "to_run.sh" in order to submit MSIsensor jobs. Run "copy_MSI_scores.sh" and "copy_MSI_sites.sh" to organize the results. Run "analysis_description_V3.py" to generate Analysis summary file

Changelog: updated paths in katmai

Versions: 1.5: Compatible with v1.4; Dockerized MSI_hg38 v1.4 pipeline to work on compute1

1.4: Compatible with v1.3; Fixed bugs for querying matched tumor-normal pairs; Added run_name for analysis description file.

1.3: Fixed minor bug (highly unusual edge cases with normal bam swaps). Recommended to rerun samples with v1.3 just in case.

1.2: Slight modification to reduce job memory requirements

1.1: Slight modification for faster parallel job submission

1.0: Pipeline developed for hg38