Data Analysis

Dr. Barsi has reached the stage of cataloguing S. purpuratus development where the Archenteron cells are next to become differentiated cells of the gut. Therefor he has collected RNA sequencing data on Archenteron cells as well as a transcriptome of the entire embryo during mid-gastrulation (30 hpf). This data (gene counts of nearly 21 000 genes) has been analysed as part of the internship as laid out in the following steps:

1 Isolate the statistically significant differentially expressed genes by comparing gene count data of the Archenteron to the transcriptome of the entire embryo.

This has been done in Rstudio using the edgeR package. Statistically significant genes

were isolated by the following conditions:

upregulated statistically significant genes = both p-value < 0.05 and log2FC > 1

downregulated statistically significant genes = both p-value < 0.05 and log2FC < -1

FC is the difference between the gene expression in GFP+ and GFP- cells and allows

for identifying which genes are differentially expression in the Archenteron. P-values

then show which of those results are statistically significant. These parameters were

calculated using the classic edgeR pairwise comparison as outlined by Chen et al.

(2020) in the edgeR user’s guide. The qCML (quantile-adjusted conditional maximum

likelihood) common and tagwise dispersion were calculated and used as inputs for the

exactTest. According to Butte et al. (2000) fold changes between 0.5-1.7 are in-

significant, so fold change thresholds of 0.5 (log2 = -1) and 2 (log2 = 1) were chosen.

2 Isolate statistically significant signalling after pairing these genes with names and

classes – genes of interest are significantly upregulated signalling genes. 2 were

found: ‘Wnt6’ and ‘Hgf Hypp_ 1329, Hypp_137, Hypp_950’.

This was done using the legacy.echinobase.org 'Quantitative Developmental

Transcriptomes of S. purpuratus' page (http://legacy.echinobase.org/shiny/quantdev/)

3 Identifying other potentially interesting signalling genes. Upregulated not statistically

4 Search the literature for to identify any previously recognised involvement of the 2

significant signalling genes were visually analysed for each of the two replicates separately. Those that appeared markedly upregulated in at least one replicate where considered potentially interesting. These genes were places into 3 categories according to their gene expression profiles which were retrieved from the legacy.echinobase.org 'Quantitative Developmental Transcriptomes of S. purpuratus' database. The genes whose expression peak or increase at 30hpf were further investigated in a literature search, whereas genes whose expression decreased at 30hpf were no longer considered for further analysis.

significant signalling genes and the other potentially significant genes in gastrulation in sea urchins. A literature search was also conducted to identify genes involved in Wnt signalling.

Barsi Lab Internship

Data Analysis