Cannot open configure file in msisensor-pro 1.2.0

[Dichromatism]

Hi,

I have tried msisensor-pro (both installed by conda and pulled from docker). When I was using hg19 as a reference and tried to make a baseline using my .bam files,

docker run pengjia1110/msisensor-pro msisensor-pro baseline -d microsatellites_new.list -i normal_2.tsv -o baseline

the software returned following errors:

`Check for the environment ...
Current work path: /
Microsatellites file: /microsatellites_new.list
Configure file path: /normal_2.tsv
Output path:/baseline
/baseline/ is not existing！ now make it! OK!
/baseline/detail is not existing！ now make it ! OK!

Load files ...
Open configure file failed, please provide valid configure file ! `

I am sure that my .bam files were indexed correctly and the names were *.sorted.dedup.bam and *.sorted.dedup.bam.bai.

When I using version 1.0.2 it returned as cannot find the index files.

So what should I do to figure this out? Thanks!

[PengJia6]

when you use docker version, you must use -v option in docker run to convert you local path to docker environment path, I will give you a small expamle.

In you computer, suppose you put your bam file at /home/user1/data/baselinebams/,
suppose you have two bam file:
/home/user1/data/baselinebams/sample1.bam and /home/user1/data/baselinebams/sample2.bam
the microsatellie list: /home/user1/data/baselinebams/microsatellites_new.list

write you configurefile (/home/user1/data/baselinebams/normal_2.tsv) as follow:

sample1     /output/sample1.bam
sample2     /output/sample1.bam

PS: the prefix /output is at the docker environment

now you can run msisensor-pro with docker as follow:

docker run -v /home/user1/data/baselinebams:/output  pengjia1110/msisensor-pro msisensor-pro baseline -d /output/microsatellites_new.list -i /output/normal_2.tsv -o /output/baseline

[Dichromatism]

Thanks to your reply!

I ran msisensor-pro successfully using pro method. The baseline were constructed using 33 normal bams. I tested it using one of my paired data. When I using msisensor msi (is this one equal to msisensor-pro msi?) to check msi sites with my normal.bam and tumor.bam, the results returned as 76 out of 196 sites were MSI. However when I using msisensor-pro pro with my tumor.bam and the baseline which I just constructed, the results returned as 5 out of 180 sites were MSI. Which of those results should I believe?

I noticed the msisensor-pro pro had different defaults in -m and -w to msisensor msi, so I changed those values but the results were same as before.

Hope to your further reply!

[PengJia6]

I thinks it may be caused by the threshold, you can try to change the -f in msi command. You can also plot a scatter diagram to see if these two results are linear correlation?