At the start of my rotation project, I was aiming to replicate Mehran's HPV integration paper results using Polyidus, the tool he developed, and get it publication-ready. However, a paper was released in 2022, benchmarking Polyidus against a new tool called Isling. Isling appeared to perform better in terms of accuracy and speed than Polyidus. So, I changed my project to instead install Isling and try it out.
Main directory: /mnt/work1/users/hoffmangroup/nhanafi
Polyidus installed here: /mnt/work1/users/hoffmangroup/nhanafi/polyidus
Isling installed here: /mnt/work1/users/hoffmangroup/nhanafi/isling
Input datasets used to be installed here (deleted fastqs to save space) + install script and sample metadata: /mnt/work1/users/hoffmangroup/nhanafi/data
Isling results found here: /mnt/work1/users/hoffmangroup/nhanafi/isling/out
"pipeline-test" refers to the in-built test that Isling provided.
"HeLa_hpv16_results" refers to a run with HeLa-S3, incorrectly using HPV16.
"HeLa-S3" refers to a run with HeLa-S3, correctly using HPV18.
"Si-Ha2" refers to a run with SiHa-2, using HPV18.
"Si-Ha3" refers to a run with SiHa-3, using HPV18.
I am not actually sure what the distinction between the 2 Si-Has are. Both were selected because they were sequenced using Illumina NextFlow, as were the RNA datasets from Mehran's paper, I believe.
sbatch script to run isling is here: /mnt/work1/users/hoffmangroup/nhanafi/isling/run_isling.sh
Isling Github: https://github.com/aehrc/isling/tree/master