LARRY Dataset: lineage and RNA recovery
Project description
This is a Python package that downloads and preprocesses the LARRY dataset from the AllonKleinLab GitHub repository. The LARRY dataset is a group of single-cell lineage-traced datasets that have been used to study transcriptional landscapes and cell fate during differentiation [1]. There are three datasets within this group of datasets, all using hematopoietic progenitor cells from mouse bone marrow and the LARRY lentiviral barcoding strategy.
- in vitro
- in vivo - transplanted into mice
- Cytokine-perturbed (in vitro) - split between various cytokine culture conditions.
The package includes functions to download the LARRY dataset, format the files into AnnData
and perform preprocessing / gene-filtering steps. The preprocessed data can then be split into test and training sets for use in different machine learning tasks. Dimension reduction can be performed after data splitting to respect information leakage or on all of the data.
Installation
pip
distribution
pip install larry-dataset
Development version
git clone https://github.com/mvinyard/LARRY-dataset.git; cd LARRY-dataset
pip install -e .
Quickstart
Downloads pre-processed data from AllonKleinLab/paper-data to ./KleinLabData
(by default). The data is formatted into AnnData
and returned to the user. A .h5ad
file is also saved, locally. The data downloading and conversion step take several minutes due to the large expression normed_counts
matrices though this only happens once.
import larry
dataset = "in_vitro" # can also choose from: "in_vivo" or "cytokine_perturbation"
adata = larry.fetch(dataset)
AnnData object with n_obs × n_vars = 130887 × 25289
obs: 'Library', 'Cell barcode', 'Time point', 'Starting population', 'Cell type annotation', 'Well', 'SPRING-x', 'SPRING-y'
var: 'gene_name'
obsm: 'X_clone'
import larry
LARRY_LightningData = larry.LARRY_LightningDataModule()
LARRY_LightningData.prepare_data()
AnnData object with n_obs × n_vars = 130887 × 25289
obs: 'Library', 'Cell barcode', 'Time point', 'Starting population', 'Cell type annotation', 'Well', 'SPRING-x', 'SPRING-y'
var: 'gene_name'
uns: 'dataset', 'h5ad_path'
obsm: 'X_clone'
Preprocessing performed previously. Loading...done.
Under the hood, the LARRY_LightningData
calls larry.fetch()
and larry.pp.Yeo2021_recipe()
, and if task == "fate_prediction"
, larry.pp.annotate_fate_test_train()
LARRY_LightningData.adata
Print the updated adata
:
AnnData object with n_obs × n_vars = 130887 × 25289
obs: 'Library', 'Cell barcode', 'Time point', 'Starting population', 'Cell type annotation', 'Well', 'SPRING-x', 'SPRING-y', 'cell_idx', 'clone_idx'
var: 'gene_name', 'highly_variable', 'corr_cell_cycle', 'pass_filter'
uns: 'dataset', 'h5ad_path', 'highly_variable_genes_idx', 'n_corr_cell_cycle', 'n_hv', 'n_mito', 'n_pass', 'n_total', 'pp_h5ad_path'
obsm: 'X_clone', 'X_pca', 'X_scaled', 'X_umap'
Sources
Repositories
References
- Weinreb, C., Rodriguez-Fraticelli, A., Camargo, F.D., Klein, A.M. Lineage tracing on transcriptional landscapes links state to fate during differentiation. Science 80 (2020). https://doi.org/10.1126/science.aaw3381
Please email Michael E. Vinyard (mvinyard@broadinstitute.org) with any questions or interests.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for LARRY_dataset-0.0.2rc0-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | a85309bff86c494a9694a209b0736d425c7b563bbd0b10d17c6289452029ef1b |
|
MD5 | aaf675dcf47fe8c29ff69665d3d393ad |
|
BLAKE2b-256 | 85ab8b1294888f25d8b4b7b4ec2d125ff9456c8890a1dcd348daac24b56ac9d9 |