SAMap comparison of A. fascicularis

Streamlined SAMap pairwise comparisons. Produced by notebooks in the samap/ folder in the https://git.embl.de/musser/chiton10x repository. The directory structure is enforced by the scripts and contains everything needed to reproduce the analysis as well as the results of the analysis. It should be possible to open the pickled files using the appropriate versions of Python and SAMap - for details please refer to the environment .yml file.

aln/:

holds the pairwise maps for all species comparisons. Four-letter species IDs were used throughout the SAMap analysis, and the folders were named accordingly (e.g. `afassmed` for the comparison of _A. fascicularis_ to _S. mediterranea_). The `species1_to_species2.txt` files include gene names that are harmonised to the corresponding `.h5ad` objects. Additionally, `/base/aln/m8/` holds the MMseqs2 output of the different comparisons in `.m8` format and the translation of the _L. vulgaris_ transcriptome (`raw/`).

raw/:

holds a folder per species. Inside every species folder are, at minimum, the original processed scRNA-seq data in `.h5` or `.h5ad` format (as provided by the authors), information about the cells (various clusterings), gene annotation (gene names, EggNOG-mapper annotation), and the proteome/transcriptome file that was used to produce the annotation.

processed/:

holds the newly made counts-and-metadata object for each species. This includes the post-QC data (only genes/cells the authors kept for their analysis) as well as clustering information and gene names, but has been reset to the raw counts.

results/:

contains the pickled SAMap objects for each comparison (for package versions please consult the conda environment `.yml` file) and a folder of results for each compared species. Each subfolder (`/base/results/species`) contains:

the pairwise cluster similarity matrix, visualized as an interactive heatmap (`all_clusters.html`) or static PDF, and saved in `.json` and `.csv` format (`similarity_table.csv`).

`gene_pairs.tsv`: a table of significant gene pairs for each cluster mapping.

`orthology.csv`: a matrix of pairwise homology for all genes in the species comparison; either 0 (no homology), 1 (same gene family) or 2 (orthologs).

the `reports/` folder, with `.html` reports for each pairwise cluster comparison above a certain similarity score threshold.