Analysis of Single-cell RNA sequencing (scRNAseq) data gives the ability to study gene expression at a cellular level. This reveals information and variation that would otherwise be masked when considering a sample as a whole, as in the case of bulk RNAseq.
scRNAseq data analyses are relevant to a range of applications such as:
- Identification of novel cell subpopulations and their gene expression profiles
- Identification of altered states of cell populations associated with drug response or survival
- Comparison of cell type composition of samples across different conditions
- Trajectory analysis to study development and differentiation of cell types and identification of genes associated with specific lineages or developmental states
- Immune receptor profiling data analysis to understand the TCR, BCR, or Immunoglobulin (Ig) clonotype composition of samples
scRNAseq data analysis of a single sample can be used to classify the cells from that sample into distinct types; this reveals information about the cell types present and how those cell types differ in terms of expression. Typically, this is done following dimension reduction and clustering using tSNE or UMAP methods. For multiple samples, scRNAseq facilitates comparisons of expression between equivalent cell types of different samples to measure the effect of a treatment, condition or outcome as well as comparisons of cell type composition of samples from multiple individuals to understand variation of cellular composition at a population scale.
What We Offer
Fios Genomics offers established, cost-efficient and rapid turnaround analysis services for single-cell gene expression data from a range of platforms including 10X Genomics, DropSeq, SMART-Seq2 etc.
We are able to receive data in various formats for analysis such as raw FastQ files, aligned BAM/SAM files, and raw count matrices.
Every time our clients work with Fios, they benefit from:
- A dedicated analyst backed by an experienced team to curate all data, identify the most appropriate statistical approach to take, and provide biological interpretation of results.
- An interactive data analysis report, internally peer-reviewed, including all analysis methods and results.
- Post-report follow-ups: upon receipt of our data analysis report, we arrange a teleconference so that our lead analyst can talk through the results.
- Access to large capacity computing and secure data storage facilities.
We have utilized the Bioinformatics team at FIOS Genomics for many of our drug discovery projects, as they provide expertise in the analysis of complex bioinformatic datasets. This includes large scale datasets from public sources as well as internally generated datasets. In many instances, at the start of a project, we have planned our large scale transcriptomic/proteomic studies with the FIOS team, to ensure that the data generated would provide the information we need, and that our projects had the highest chance of success. We have been consistently impressed with the rigor of FIOS’ work, their communication throughout the projects, and the rapid speed at which they complete their analyses.
Every time our clients work with us, they benefit from:
- A dedicated analyst backed by an experienced team to curate all data, identify the most appropriate statistical approach to take and provide a biological interpretation of results.
- An interactive data analysis report, internally peer-reviewed, including all analysis methods and results.
- Post-report follow ups: upon receipt of our data analysis report, we arrange a teleconference so that our lead analyst can talk through the results.
Access to large capacity computing and secure data storage facilities.
An example analysis pipeline for single-cell data includes:
- Pre-processing, alignment and/or quantification using standard bioinformatics tools (e.g. STAR, CellRanger, salmon, kallisto).
- Quality control evaluation of raw expression data, including doublet detection (e.g. scrublet).
- Batch effect correction and alignment for combined analysis of multiple datasets (e.g. Seurat CCA, Harmony, LIGER).
- Unsupervised clustering of the data to identify cell populations using graph-based clustering algorithms and visualisation with tSNE or UMAP.
- Quality control evaluation of identified clusters to eliminate technical or irrelevant biological (e.g. cell cycle) bias.
- Cell type annotation of identified clusters.
- Differential expression analysis.
- Functional enrichment analysis using curated resources such as the Reactome pathway database and the Gene Ontology (GO) resource.
- Pseudotemporal ordering of cells to identify developmental trajectories (e.g. monocle, TSCAN).
- Inference of cell-cell interactions between clusters (e.g. CellPhoneDB).
What we have helped with
Below is a small selection of projects where we have successfully helped our clients:
- Single-cell RNAseq analysis of a publicly available dataset, where normal and malignant epithelial cells were identified from transcriptome profiles of ~200,000 single cells from non-small cell lung cancer (NSCLC) patients.
- The evaluation of gene expression changes in response to treatment in alpha, beta, delta and ductal cells from pancreas tissue samples.
- Association of gene expression and functional pathways with immune cell subtypes in peripheral blood mononuclear cells (PBMCs).
We offer a wide range of services:
Discovery
Selecting the correct targets and/or the correct indication is essential for development success. We help support the process with robust analysis of historic or new data.
Preclinical Research
We help in experimental design and statistical analysis and guide our clients in making informed decisions during the preclinical stage.
Clinical
Research
We offer a comprehensive analysis approach for augmenting clinical trial outcomes, ensuring you get the most information out of your research.
Drug Repurposing
We have strong experience with identifying in silico new potential indications for existing drugs, reducing the cost and time of downstream wet lab validation.
