3 Configuration
Once you have prepared a database, you need to point scMarco to it by editing a configuration file (config.yaml) that you can find in the directory.
The configuration file would contain several parts which we will discuss in this section.
Each part is defined by a name and a value separated by a colon( e.g., name: value), and indentation is used to show hierarchy. For example, example_optic_lobe sits under and belongs to databases, while flybase_tf is a part of gene_list.
title: "scMarco"
database_path: 'example/min_db.sqlite'
databases:
example:
label: "Subset of optic lobe atlas from Ozel et al. (2021)"
stages: ['P15', 'P30', 'P40', 'P50', 'P70', 'Adult']
gene_list:
flybase_tf:
path: "data/flybase_tfs.txt"
label: "FlyBase TFs"
mimic:
path: "data/mimic_temp.txt"
label: "RMCE swappable gene-specific lines"3.1 title
This defines the title that you will see when you run scMarco.
3.2 database_path
This is the file path to the database you generated when you preprocessed your dataset. The path should be relative to the home directory of the app (e.g., if you have scMarko on your desktop (~/Desktop/scmarko/) and a database (db.sqlite) is stored in a subdirectory named my_dbs, you should set database_path to my_dbs/db.sqlite).
3.3 databases
Under this part, each item (e.g., example) is a dataset. The items are the table names you set when you preprocessed your dataset.
label defines how this dataset is referred to in the app. stages defines the order of stages 1.
3.4 gene_list
There are tens of thousands of genes in a dataset, and oftentimes we are only interested in some of them. Only keeping the subset of genes helps with the efficiently of the app. You may curate and provide lists of gene symbols as txt files in which each line is a gene symbol.
In config.yaml, you give a short name (e.g., flybase_tf and mimic) to your list, and provide the relative path in path and how the list is referred to in the app in label.