evotec_Nov23_PanOmicsDriven - 38

The PanHunter Platform
Identifying and linking molecular markers unambiguously to
a specific type of cancer is a crucial achievement towards
better and affordable point-of-care diagnostics.
a 2D plot of the dimensionality
reduction (here, using the t-SNE
algorithm) of the molecular
expression data (Figure 7).
In this plot, samples with
similar gene activity are located
close to each other, while those
with larger differences are located
further apart.
As a result, a strong correlation
of the tissue definition
with the two major clusters was
immediately visible, indicating
a significant difference in gene
activity between both tissue
types. To focus on tumorous
samples only, non-tumours
samples can be deselected and/
or the dataset reduced by free
selection.
A new 2D plot, only using
the new sample subset, was
generated immediately, showing
an evenly distributed cloud of
samples without any obvious
clusters (not shown). To assist
Figure 8: Result from exploratory analysis, listing several genes which might
be driving the distribution of samples within the 2D plot of Figure 7.
ENSG0000013419J
15
5
10
Figure 9: t-SNE plot coloured according to transcription intensity of the REG4 gene.
t-SNE component 1
38
GENengnews.com
t-SNE component 2

https://www.genengnews.com https://www.genengnews.com

evotec_Nov23_PanOmicsDriven

Table of Contents for the Digital Edition of evotec_Nov23_PanOmicsDriven