#Tutorials
##scLVM scLVM is implemented in python and can either be run it in python as demonstrated in the demo ipython notebook or as an R package as demonstrated here.
In case you would like to distribute the computations over many cores, we provide scripts for easy parallelisation here: scripts
##Input Our software requires as input a count table with read counts for all cells. This can be generated e.g. with HTSeq and appropriate filtering and low-level processing should be perfromed to filter out spurious/'bad' cells before the data are analyzed within scLVM.
Our R package has convenience functions for processing this count table and the subsequent analysis with scLVM.
If you would like to use the python implementation, you can use our R scripts to process this filtered count table, both when spike-ins are present and for data-sets without spike-ins:
A demo script which can be run for data where spike-ins are present can be found here: transform_counts_demo.Rmd
Without spike-ins, baseline variability, which is required by scLVM, can be estimated as shown here: transform_counts_demo_no_spikeins.Rmd
These R scripts generate an hdf5 file containng the variables required by the core scLVM algorithm:
- Normalised gene expression data -
- Technical noise (in log space)
- Gene symbols
- Heterogeneous genes (boolean vector)
- Cell cycle genes (vector of indices)