1. Computer Science & Engineering Department University of Connecticut
SC1 : Computational Methods for the Analysis of Single Cell RNA-Seq Data
Marmar Moussa
Ion Măndoiu
Computer Science & Engineering Department University of Connecticut

2. Outline Motivation TF-IDF based Gene Selection & Clustering
Workflow overview QC
Gene Selection
Clustering
Differential Expression Analysis
Enrichment Analysis
Visualizing cells , gene expression, gene pairs, DE analysis, clustering heatmaps

3. Motivation No widely adopted analysis workflow yet
An interactive web-based pipeline for scRNA-seq data analysis that allows researchers to analyze data in an efficient work-flow.
Only few tools that allow researchers to analyze scRNA-seq data without considerable coding efforts.
Provide support for more detailed analysis & new methods:
several quality control (QC) options
TF-IDF based gene selection & clustering
LSImpute
Cell Cycle Analysis
Interactive choices, 3D visualization, informative plots …

4. scRNA-Seq Analysis Workflow
Data Exploration& QC
Imputation
Normalization/ Batch Effect
Data Transformation
Log Transform
TF-IDF
Feature Selection
Markers Genes
TF-IDF Genes
Dimensionality Reduction
PCA
t-SNE …
Optimal Number of Clusters, Within Cluster Homogeneity
Clustering
similarity…
DE Analysis
Enrichment
Analysis,
Annotation
Visualization
Cell Cycle Analysis
Many methods available
K-means
Hierarchical clustering
Graph based …
Active area of research
Reducing effect of confounders such as detection rate & cell cycle phase
Discriminative similarity metrics
Scalability to millions of cells…

5. QC Dashboard

6. Gene Selection & Clustering Single Cell RNA-Seq Data using TF-IDF based Methods
Goal  clustering cells into cell types …
Moussa, M., and Mandoiu, I.: Clustering scRNA-Seq Data using TF-IDF. Bioinformatics Research and Applications. 13th International Symposium, ISBRA 2017, Honolulu, HI, USA. Lecture Notes in Computer Science book series ( Extended Abstract in LNCS, volume 10330).
Moussa, M., and Mandoiu, I.: Single Cell RNA-seq Data Clustering using TF-IDF based Methods. (BMC Genomics, 2018).

7. TF-IDF Transformation
Term Frequency x Inverse Document Frequency
Successfully employed in information retrieval
Two parts:
How many times a term occurs in a document? (Considers term frequency )
How ‘important’ the term is?(Considers document/collection frequency )
(Intuition: rare terms in a collection are more informative than frequent terms; Think stop-words!)
Topics classification

8. TF-IDF Transformation
Term Frequency x Inverse Document Frequency for scRNA-Seq data:
For gene i in cell j with count f :
𝑇 𝐹 𝑖𝑗 = 𝑓 𝑖𝑗 / max 𝑘 𝑓 𝑘𝑗
If gene i is detected in ni out of N cells:
𝐼𝐷 𝐹 𝑖 = log 2 (𝑁/ 𝑛 𝑖 )
TF-IDF score:
𝑇 𝐹 𝑖𝑗 ∗𝐼𝐷 𝐹 𝑖
Analogy to scRNA-seq data using UMIs
Library/collection/corpus of documents – sample of cells
Document – cell
Term -- Genes

9. TF-IDF based Feature selection
Genes with highest avg. TF-IDF (highest mean GMM in red)

10. scRNA-Seq Clustering Methods
Cells QC, Genes QC, Gap-Statistics Analysis
Data Transformation: Log2(x+1) or none
Feature Selection: PCA, tSNE, highly variable genes* or none
Seurat (K-means)*
Seurat (SNN)*
GMM
K-means
Sph. K-means
HC (E/P)
Louvain (E)
Data Transformation:
TF-IDF
Feature Selection: High avg. TFIDF score (Top) or Highly variable TF-IDF (Var)
HC (E/P/C)
Data Binarization:
Cutoff threshold per cell based on cell avg. TF-IDF(Bin)
HC (E/P/C/J)
Greedy (E/P/C/J)
Louvain (E/P/C/J)

11. Pairs by ‘difficulty’: dissimilar  intermediate  similar
highly dissimilar: (b cells and cd14 monocytes) and (b cells and cd56 nk)
highly similar : (memory t and naive cytotoxic) and (regulatory t and naive t)
intermediate similarity: (memory t and naive t) and (regulatory t and naive cytotoxic)

12. PBMC, 10x 5: Monocytes, 6: Natural Killer 4: B cells
7,8: naïve cytotoxic, cytotoxic, activated cytotoxic
1: helper
2: regulatory

13. DRG-neurons, C1
42 Genes out of 47 hand-curated Markers List picked in top TF-IDF Genes

14. T-cells, 10x

16. Future Work
Support for additional clustering & DE analysis methods, etc.
Integrate imputation
Moussa, M., Mandoiu, I.: Locality sensitive imputation for single-cell rna-seq data. ISBRA2018 Proceedings (to appear in ISBRA 2018 proceedings) . bioRxiv preprint available at doi:  (
Cell Cycle Analysis

17. Demo

18. Thank You.
Questions?

19. TF-IDF based gene selection
Genes with highest avg. TF-IDF, ranked (Top):
Density
fitted n-mixture GMM model to the distribution of TF-IDF gene averages, ranked by number of detecting cells
Heavy tail
Genes with highest avg. TF-IDF (highest mean GMM in red)
Avg. Gene TF-IDF score for regulatory, memory cells mix