dpeerlab / sckinetics

Code for scKINETICS (ISMB 2023)

scKINETICS

We introduce scKINETICS (Key regulatory Interaction NETwork for Inferring Cell Speed), a dynamical model of gene expression change which is fit with the simultaneous learning of per-cell transcriptional velocities and a governing gene regulatory network. This is accomplished through an expectation-maximization approach derived to learn the impact of each regulator on its target genes, leveraging biologically-motivated priors from epigenetic data, gene-gene co-expression, and constraints on cells’ future states imposed by the phenotypic manifold.

Demo

The demo shows the main functionalities of the package including input file generation, GRN construction, EM estimation, visualizations, and the TF ablation experiment.

Paper

https://academic.oup.com/bioinformatics/article/39/Supplement_1/i394/7210448

sckinetics's Issues

Indexing into list of genes with a covariance matrix

Hi! I'm a causal inference and transcriptional data enthusiast, and I'm excited to try out scKINETICS because I think the backronym is really cool and it has a really neat way of getting around the steady state assumption that Dictys and CellOracle make. I ran into an error on this line.

prior_genes[celltype] = np.array([columns.index(gene) for gene in list(celltype_priors[celltype])])

It seems to:

• make a list containing a single covariance matrix, or a chunk of one.
• iterate over that list, storing the matrix in a variable gene.
• index into columns.index looking for the integer position of gene.

Should gene not be a string containing a gene name, or an int? Why is it not the following?

prior_genes[celltype] = np.array([adata.var_names.index(gene) for gene in adata.var_names])

Thank you!

some dependencies missing in requirements.yml

Dear developers, thank your for sharing your work, you did an amazing job.

I had some issues during the setup of the conda environment because of some conflicts between the R version installed within conda environment and another installation that I already had (basically the problem is caused by rpy2 and I could solve it setting the right LD_LIBRARY_PATH in the environment variables, etc...)

Then I had to install some missing dependencies (TxDb.Mmusculus.UCSC.mm10.knownGene, genomeinfodb; genomeinfodbdata; chipseeker). It would be great if they were already specified in the requirements file.

Thanks!

Lídia

Integrating scRNA-seq, scATAC-seq, and bulk ATAC-seq

Hello,

Great tool. We are interested in using scKINETICS for an experiment in which we have scRNA-seq, scATAC-seq, and bulk ATAC-seq data. Is it possible to use all three modalities within the scKINETICS framework? And if so are there any demonstrations available?

Best,
Bryan

Documentation availability

Hi! I really enjoyed reading the paper and I would like to know when the documentation webpage https://sckinetics.readthedocs.io/en/latest/index.html will be updated.
Thank you very much in advance for your answer!