scLiTr is a Python package for analysis of lineage tracing coupled with single-cell RNA-Seq.
The main key of the package are clonal embeddings — vector representations of the whole clones in low dimensional space (clone2vec). These representations is a dropout-robust and cluster-free way of representation of heterogeneity within clonal behaviour for cell type tree-free hypothesis generation regarding cells' multipotency.
clone2vec builds representation of clones in exact same way with popular word embedding algorithm — word2vec — via construction two-layers fully connected neural network (it uses Skip-Gram architecture) that aims to predict neighbour cells clonal labellings by clonal label of cells. As a result, clones that exist in similar context in gene expression space will have similar weights in this neural network, and these weights will be used as embedding for further analysis.
scLiTr might be installed via pip
(takes 1-2 minutes on Google Colab):
pip install sclitr
or the latest development version can be installed from GitHub using:
pip install git+https://github.com/kharchenkolab/scLiTr
scLiTr
requires Python 3.8 or later with packages listed in setup.cfg file. The package was successfully tested
on the following systems:
- macOS Sonoma 14.5 (Apple M1 Chip @ 3.20GHz × 8, 16GB RAM) — MacBook Air M1,
- Ubuntu 18.04.5 LTS, 64-bit (Intel Xeon @ 2.60GHz × 32, 256GB RAM) — PowerEdge server,
- Ubuntu 22.04.3 LTS, 64-bit (Intel Xeon @ 2.20GHz × 2, 13GB RAM) — Google Colab.
Please visit documentation web-site to check out API description and a few tutorials with analysis.
An example with the dataset from Weinreb et al., 2020 is available on Google Colab (takes about 45 minutes on the CPU, of which approximately 30 minutes are clone2vec latent representation construction).
For interactive exploration of clonal and gene expression embeddings together we recommend using our simple tool clones2cells.