Description of the problem

• What mechanisms underlie the disease?
• What parts of the data were affected the most?
• How these differences can be described?
• What do these differences mean biologically?

Existing solutions

Align samples

scVI, Conos, ..., Seurat

See the review from the Theis lab

*Integrating single-cell transcriptomic data across different conditions, technologies, and species, Butler A, et al. Nature Biotechnology, 2018

Existing solutions

Align samples

Perform joint annotation

*Integrating single-cell transcriptomic data across different conditions, technologies, and species, Butler A, et al. Nature Biotechnology, 2018

Existing solutions

Align samples

Perform joint annotation

Run differential expression 

*Single-cell genomics identifies cell type–specific molecular changes in autism, Velmeshev D, et al. Science, 2019

Existing solutions

Align samples

Perform joint annotation

Run differential expression 

Run Gene Ontology analysis

*Single-cell genomics identifies cell type–specific molecular changes in autism, Velmeshev D, et al. Science, 2019

Existing solutions

Align samples

Perform joint annotation

Run differential expression 

Run Gene Ontology analysis

Compare cell type proportions

*Schirmer, L., Velmeshev, D., Holmqvist, S. et al. Neuronal vulnerability and multilineage diversity in multiple sclerosis. Nature 573, 75–82 (2019).

Existing solutions

Align samples

Perform joint annotation

Run differential expression 

Run Gene Ontology analysis

• A single-cell atlas of entorhinal cortex from individuals with Alzheimer’s disease reveals cell-type-specific gene expression regulation, Grubman A, Nature Neuroscience (2019).

• Single-cell genomics identifies cell type–specific molecular changes in autism, Velmeshev D, et al. Science (2019).

• Mathys, H., Davila-Velderrain, J., Peng, Z. et al. Single-cell transcriptomic analysis of Alzheimer’s disease. Nature 570, 332–337 (2019).

• Schirmer, L., Velmeshev, D., Holmqvist, S. et al. Neuronal vulnerability and multilineage diversity in multiple sclerosis. Nature 573, 75–82 (2019).

• Wheeler, M.A., Clark, I.C., Tjon, E.C. et al. MAFG-driven astrocytes promote CNS inflammation. Nature 578, 593–599 (2020).

Compare cell type proportions

What can we possibly do?

Compositional analysis

Cluster-based

Cluster-free

Case-control analysis of single-cell studies: a fresh approach

Composition analysis, cluster-based

Problem: changes are not independent

*Credit to Anna Igolkina

Composition analysis, cluster-based

Problem: changes are not independent

*Credit to Anna Igolkina

Composition analysis, cluster-based

*Schirmer, L., Velmeshev, D., Holmqvist, S. et al. Neuronal vulnerability and multilineage diversity in multiple sclerosis. Nature 573, 75–82 (2019)

Composition analysis, cluster-based

*Schirmer, L., Velmeshev, D., Holmqvist, S. et al. Neuronal vulnerability and multilineage diversity in multiple sclerosis. Nature 573, 75–82 (2019)

Composition analysis, cluster-free

*Schirmer, L., Velmeshev, D., Holmqvist, S. et al. Neuronal vulnerability and multilineage diversity in multiple sclerosis. Nature 573, 75–82 (2019)

Effect size

Significance

Expression analysis, cluster-based

Visualization of sample structure

*Schirmer, L., Velmeshev, D., Holmqvist, S. et al. Neuronal vulnerability and multilineage diversity in multiple sclerosis. Nature 573, 75–82 (2019)

Expression analysis, cluster-based

Visualization of sample structure

*Schirmer, L., Velmeshev, D., Holmqvist, S. et al. Neuronal vulnerability and multilineage diversity in multiple sclerosis. Nature 573, 75–82 (2019)

Expression analysis, cluster-based

How separated our conditions are?

*Schirmer, L., Velmeshev, D., Holmqvist, S. et al. Neuronal vulnerability and multilineage diversity in multiple sclerosis. Nature 573, 75–82 (2019)

Expression analysis, cluster-free

Differential expression on single cells

Expression analysis, cluster-free

Differential expression on single cells

Expression analysis, cluster-free

Differential expression on single cells

Expression analysis, cluster-free

Differential expression on single cells

Expression analysis, cluster-free

Gene programs

*Schirmer, L., Velmeshev, D., Holmqvist, S. et al. Neuronal vulnerability and multilineage diversity in multiple sclerosis. Nature 573, 75–82 (2019)

Full analysis of the Epilepsy dataset

Compositional analysis

Full analysis of the Epilepsy dataset

Compositional analysis, cluster-free

Full analysis of the Epilepsy dataset

Compositional analysis, cluster-free

Effect size

Significance

Full analysis of the Epilepsy dataset

Expression analysis: sample embeddings

Condition

Full analysis of the Epilepsy dataset

Expression analysis: sample embeddings

Condition

Sex

Full analysis of the Epilepsy dataset

Expression analysis: sample embeddings

Condition

Sex

Protocol

Full analysis of the Epilepsy dataset

Expression analysis: shift magnitudes

Full analysis of the Epilepsy dataset

Expression analysis: shift magnitudes

Full analysis of the Epilepsy dataset

Expression analysis: shift magnitudes

Full analysis of the Epilepsy dataset

Programs for excitatory neurons

Full analysis of the Epilepsy dataset

Global excitatory program

Full analysis of the Epilepsy dataset

Programs for inhibitory neurons

Full analysis of the Epilepsy dataset

Joint program

Full analysis of the Epilepsy dataset

Expression analysis: local cluster-free DE

Full analysis of the Epilepsy dataset

Expression analysis: local cluster-free DE

Full analysis of the Epilepsy dataset

DE instability

Use top DE genes instead of p-cutoff!

Epilepsy dataset

Cancer dataset

Full analysis of the Epilepsy dataset

DE instability

Use pseudo-bulk DE, not single-cell methods!

Squair JW, et al. Confronting false discoveries in single-cell differential expression

Zimmerman KD, et al. A practical solution to pseudoreplication bias in single-cell studies

Single-cell-based methods fail even on a theoretical level