PANORAMICS - A Vision
Re-imagining conversations in science
a pan-Canadian working cluster for spatial & single cell multi-omics analyses
The "PANORAMICS TUTORIAL-Lecture series " for SINGLE-CELL RNA SEQuencing
Chapter 4 | Let's deviate from tradition
Dimensional reduction methods are instrumental in single cell analysis and in data visualization. However, dimensional reduction does not stop at data "shrinkage" nor simply serve as a pre-processing step prior to application of graph-based methods or to facilitate visualization. How often do we question the our choice of standard methods applied in dimensional reduction? Are there alternative methods that we can apply?
Similarly integration tools have become frequently used in data assessment and processing. And yet, standard integration methods faces challenges when faced with complex datasets, such as cancer datasets. Perhaps it is time to explore newer methods developed from AI.
Here in our next lecture series under our "challenge-me" series, we discuss non-traditional methods of dimensional reduction in data analysis. In addition, with the rise of AI in method developments, we introduce here a topic on deep-learning application in data integration.
In this discussion we will address some common challenges when performing differential expression analysis from scRNAseq data. These will include the introduction of biases due to selective inference (aka "double-dipping", calculating differential expression between groups defined by differences in expression) and pseudoreplication (it is rarely correct to consider each cell a replicate). We will discuss the scientific literature around avoiding or accommodating for these biases.
Mr. Brendan Innes
PhD Candidate
(The Bader Lab)
To learn more about the Brendan Innes, click here
4th July 2023 | 1100-1200hrs
new revised timing & date
a "challenge-me" series event (LECTURE)
High-dimensional analysis - Thinking beyond tradition
Miss. Delaram Pouyabahar
Bader Lab
click to learn more
Single-cell RNA-sequencing is able to identify the gene expression heterogeneity within complex biological systems, though interpretation is challenging due to a mix of biological and technical factors. Previous studies have demonstrated the utility of reduced dimensional representations to identify shared cellular attributes and unique biological processes across single-cell datasets.However, in many cases, the inferred dimensions from standard matrix factorization methods may not align with biologically meaningful gene expression programs, and nonlinear methods often lack interpretability. In this session, we’ll discuss factor-analysis-based approaches for identifying and interpreting the biological and technical sources of variation in single-cell transcriptomics maps. We’ll demonstrate the application of an effective factor analysis model, Varimax PCA, on a novel single-cell map of the healthy rat liver, providing a good example of the utility of factor decomposition for single-cell genomics data analysis and interpretation.
5th July 2023 | 1200-1300hrs
a "challenge-me" series event (LECTURE)
pan-cancer scRNAseq data integration at scale using deep-learning
