1 Data Settings

CROP-seq data are from this paper: High-throughput single-cell functional elucidation of neurodevelopmental disease-associated genes reveals convergent mechanisms altering neuronal differentiation, GEO accession: GSE142078.

Perturbations:
CRISPR knock-down of 14 autism spectrum disorder (ASD)–associated genes (3 gRNAs per gene) + 5 non-targeting gRNAs.

Cells:
Lund human mesencephalic (LUHMES) neural progenitor cell line.
Cells were sequenced in 3 batches.

2 scRNA-seq data

2.1 Preprocessing

There are 8708 cells with a single type of gRNA readout using the quality control criteria of % mitochondria gene expression < 10 and total UMI count < 2e+4.

Only genes detected in > 10% of cells were kept.

Seurat "LogNormalize": log(count per \(10^4\) + 1).

Batch effect, (unique UMI count,) library size, and mitochondria percentage were all corrected for.

2.2 UMAP representation after correction

Batch effect:

No batch effects were apparent on global clustering after correction.

Cell cycle effect (colored by the mean expression of signature genes at the corresponding cell cycle stage):

Expression patterns of neuronal marker genes:

Most cells were postmitotic as assessed by the absence of proliferation marker TOP2A.
But there is heterogeneity in the progression of differentiation at the single-cell level.

3 CRISPR Perturbations

3.1 Distribution of Knock-down Perturbations

3.2 DGE -- Wilcoxon rank sum test on expression without correction

Number of genes that passed FDR < 0.05 under each perturbation:

KO	ADNP	ARID1B	ASH1L	CHD2	CHD8
DE_genes	30	132	56	161	100
KO	CTNND2	DYRK1A	HDAC5	MECP2	MYT1L
DE_genes	0	0	2	1	0
KO	Nontargeting	POGZ	PTEN	RELN	SETD5
DE_genes	0	0	523	0	5

3.3 DGE -- Welch's t-test on corrected expression data