New Functionality: Find Marker Genes
Embed new cells in an already existing space
GFICF package try to identify marker genes across clusters performing Mann-Whitney U test with continuity correction. Briefly differentially expressed (DE) genes of each cluster are identified comparing the expression of each gene in each cluster versus the all the others. See the example here about t...
RcppParallel Mann–Whitney U test
Embed new cells in an already existing space
I have implemented an RcppParalle and Rcpp version of Mann–Whitney U test with continuity correction, that using 6 threads was about 100X faster then native R function and 5X faster then corresponding serial C++ function. Test were performed using i7 with 6 cores.
The implemented C++ Mann–Whitney U t...
New Functionality: Predict active pathways
Embed new cells in an already existing space
We can now use GSEA (Gene Set Enrichement Analysis) to predict pathway activity across cells of the same cluster. Briefly gf-icf gene ranks are first summedd across cells of the same cluster and then GSEA is performed. This analysis can be now performed via runGSEA function. See the example here about...
Parallel Jaccard with RcppParallel
Parallel Jaccard Benchmarks
We have implemented a naive but fast parallel version of Jaccard Coefficient estimation for the Phenograph method, thus obtaing a speed boot of about 20X when comparade with previous serial implementation.
Tets were performed using a dataset composed by 32900 cells and using different numbers of neighbors for each ce...
New Functionality: Emmbed new cells
Embed new cells in an already existing space
You can now use an existing embedding to add new cells via embedNewCells. Briefly new cells are first normalized with GF-ICF method but using as ICF weigth estimated on the training set and than projected in the existing PCA/LSA space before to be embedded in the already existing UMAP space via uwot::...
GF-ICF pubblished
Abstract
Gene expression in individual cells can now be measured for thousands of cells in a single experiment thanks to innovative sample-preparation and sequencing technologies. State-of-the-art computational pipelines for single-cell RNA-sequencing data, however, still employ computational methods that were developed for traditional bulk RNA-...