gdagstn / cellula

Simple implementation of different single cell RNA-seq methods

Hi there 👋

I am a computational biologist - originally trained as molecular biologist - who started learning data analysis in R in 2017. My current areas of expertise include transcriptomics (bulk and single cell), translatomics, neurodevelopmental disorders, neurodegenerative diseases, genome editing, and stem cell biology. I am particularly passionate about data visualization.

Since early 2021 I started honing my skills as a package developer. So fare I have released seven R packages in the wild:

• 🧠 {coldcuts} - draw and render neuroanatomical segmentations in 2D and 3D
• 🍳 {oveRlay} - annotate 2D scatterplots with overlay polygons
• 🎨 {swatcheR} - generate color swatches from pictures
• 🦜 {alamak} - let Pixel Pals handle your errors
• 🗺️ {borges} - antique cartographic style 2D point clouds (UMAP, tSNE and the like)
• 🧫 {cellula} - pipeline with wrappers to many single cell RNA-seq analysis methods
• 🏢 {HDB} - a geometric test to detect batch effects in single cell RNA-seq

with hopefully more to come.

🛠 I am currently learning three.js, a little bit of Python on the side, and Arduino physical computing as a hobby.

cellula's Issues

When using assigning identities, some cell annotation labels are missing for the UCell method

When using the UCell method of cell annotation with assignIdentities(), some of the genes are missing. The AUC method seems to include all of the gene sets when the exact same SCE object and geneset is used. A warning message is given as shown below.

`[ANNO/UCell] Calculating UCell scores
Warning: The following genes were not found and will be
imputed to exp=0:

• CLEC4OP,RYR3-DT,DYNLT5,ENSG00000225792,ENSG00000226780,ENSG00000229425,ENSG00000234432,H2AW,ITGB8-AS1,UBR5-DT,ENSG00000233968,CDK6-AS1,CFAP251,ENSG00000263667,ENSG00000264269,H1-2,H2BC8,AQP5-AS1,DNMT3L-AS1,ENSG00000227220,ENSG00000236166,LOC100506688,RFX5-AS1,ZNF22-AS1,ENSG00000230537,CRYM-AS1,LCT-AS1,LOC105377488,NOPCHAP1,PAPLN-AS1,ATP5MK,PGAP4,H1-10,H1-4,H2AJ,RNASE2CP,AARS1,BSG-AS1,CRACD,DYNC2I2,ENSG00000189316,EPRS1,EPS15-AS1,GARS1,GFUS,H1-0,H1-3,HARS1,IARS1,IFTAP,LARS1,MACIR,METTL13,MIX23,NARS1,POLR1G,RARS1,SLC30A6-DT,SLX9,TARS1,VARS1,ZNF788P,ENSG00000230709,MARCHF1,PLEKHG7,CCL3L3,CYRIA,H3-3B,PELATON,PTPRN2-AS1,WARS1,ANKRD44-AS1,CARS1,CUTALP,DCTN6-DT,PARP11-AS1,PP7080,PRKCH-AS1,RAB33B-AS1,ATP5MJ,ENSG00000238005,ENSG00000238142,GPX1,MIR223HG,MIR23AHG,MTARC1,CYRIB,ENSG00000167912,ENSG00000265206,IRF1-AS1,KIR3DX1,LINC00869,LINC02481,TFAP2E-AS1,CIBAR2,DTNB-AS1,ENSG00000249679,ENSG00000253633,ENSG [... truncated]`

the 'ssGSEA' method of cell annotation is giving an error

When using the assignIdentities() function with method = "ssGSEA", I receive an error as shown below. The exact same single cell experiment object and geneset was used with two other methods (AUC and UCell) and no error was given. New columns with scores were successfully added to colData(sce) for those methods.

# ssGSEA
sce = assignIdentities(sce, genesets = all_HAY_sets, 
                               method = "ssGSEA")

Error in .local(expr, gset.idx.list, ...) : The 'annotation' argument must contain a character string. In addition: Warning message: In .local(expr, gset.idx.list, ...) : Using 'SingleCellExperiment' objects as input is still in an experimental stage.