Project 2
Copy number alterations, where a whole segment of chromosome is duplicated are deleted, form one of the main types of genomic instability responsible for carcinogenesis. Although CNV detection was traditionally performed on whole-genome sequencing data, CNV calling methods for single-cell and spatial transcriptomic data provide a cell-level resolution of CNVs, allowing to identify tumor subclones and their genealogy and reconstruct spatial patterns of DNA damage. Among these methods, novel methods allow to perform allele-specific calls, providing information about maternal and paternal chromosome copies and precise copy number states.
In this project, you will apply a newly developed method (Xclone, https://github.com/single-cell-genetics/XClone) on malignant pleural mesothelioma single-cell data obtain in the context of the MESOMICS project (https://rarecancersgenomics.com/mesomics/). Mesothelioma is a rare and deadly disease often triggered by copy number alterations affecting key genes (see Mangiante et al. Nat Genet 2023 and its short research briefing).
Fig. 1 | The XClone pipeline. From https://xclone-cnv.readthedocs.io/en/latest/about.html
- test data from the XClone module Glioma tumors (e.g., https://xclone-cnv.readthedocs.io/en/latest/BCH869_XClone_tutorials.html). Data in the python module, e.g., accessible with xclone.data.bch869_rdr() and xclone.data.bch869_baf()
- single cell alignment data (bam file), expression data (count matrix), and annotated data (anndata object). Available on osiris at /data/Training-MG/files/data/Project2_CNVcalling_mesothelioma
- bash and python scripting
- Install the module and all dependencies following instructions on https://github.com/single-cell-genetics/XClone
- Run software on demo dataset from the XClone module (e.g., the glioma sample)
- Run preprocessing of mesothelioma sample following instructions (https://xclone-cnv.readthedocs.io/en/latest/preprocessing.html) to generate the inputs to the XClone method
- Run the XClone method on the IARC HPC (SLURM scheduler, see with supervisor for further instructions http://collab.iarc.fr/communities/ScientificComputing/_layouts/15/start.aspx#/SitePages/Community%20Home.aspx)
- Interpret the results. Compare the detected CNVs to that reported in Mangiante et al. Fig. 3.
- data preparation (will require data wrangling skills)
- python coding
- using HPC
- see software documentation https://xclone-cnv.readthedocs.io/en/latest/#
- mesomics project information https://rarecancersgenomics.com/mesomics/