Skip to content

pritykinlab/gloss-paper-analysis

Repository files navigation

gloss-paper-analysis

Companion to Gloss package, holding paper analysis. Source package: https://github.com/pritykinlab/Gloss

Folder contents are as listed below:

  • gut_experiments: scripts for replicating results on running Gloss and baseline models on original and perturbed data from the uLIPSTIC Gut dataset.
  • lcmv_experiments: scripts for replicating results on running Gloss and baseline models on original and perturbed data from the uLIPSTIC LCMV LN and Sys datasets.
  • lipstic__tumor_experiments: scripts for replicating results on running Gloss and baseline models on original and perturbed data from the LIPSTICv1 dataset.
  • perturbation_simulations: notebooks used to generate semi-simulated perturbations on the uLIPSTIC datasets.
  • prep_and_process_data: notebooks used for initial loading and preprocessing of LIPSTICv1 and uLIPSTIC datasets and their subsets.
  • geneset_curation: notebooks and scripts used for loading and processing both pathways for training Gloss models and genesets used for testing enrichments.
  • compute_enrichments: notebooks and scripts used for computing and plotting geneset enrichments for original and perturbed datasets.
  • performance_plots: contains notebooks and scripts for producing plots concerning comparisons between Gloss and baselines.
  • cite_seq_experiments: contains notebooks and scripts for producing results concerning applying Gloss to CITE-seq data on the 10K human PBMC dataset.
  • transfer_experiments: contains noteboks and scripts for producing results concerning applying trained Gloss models to matched LIPSTIC-free data from Ishizuka et al.
  • figure_plots: contains notebooks and scripts for producing manuscript figures that don't concern enrichments, performance, the transfer data, or the CITE-seq data.

Datasets validated over can be found as described in

Nakandakari-Higa, S. et al. Universal recording of immune cell interactions in vivo. Nature 627, 399–406 (2024). https://doi.org/10.1038/s41586-024-07134-4

and in

Chudnovskiy, A. et al. Proximity-dependent labeling identifies dendritic cells that drive the tumor-specific CD4+ T cell response. Sci. Immunol. 9,eadq8843 (2024). https://doi.org/10.1126/sciimmunol.adq8843

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages