This page displays the aBSREL likelyhood ratio test (LRT) results for the selected gene. Every row represents an aBSREL LRT performed for a specific phylogenetic branch (i.e. species). The Homo sapiens, A. cahirinus and M. musculus branches were tested for positive selection in every analyzed gene, while for some genes an additional random selection of 3 rodent branches and 9 non-rodent branches were tested for positive selection. Nevertheless, only the A. cahirinus branch LRT results were utilized for the generation of the A. cahirinus positively selected genelist.

This page displays the STRING clusters that were identified in the A. cahirinus (episodic) positively selected ABSREL genelist. The clusters were generated using the Markov cluster (MCL) algorithm with an inflation parameter of 4

We identified 8 STRING clusters with at least 5 genes and 64 STRING clusters with the minimum of 2 genes These can be explored using the interactive scatterplot and table below. Furthermore, the desired STRING clusters in the table can be selected using the range buttons on the left side of the page.

Scatterplot of A. cahirinus P-value versus non-synonymous/synonymous substitution rate

Double click a cluster in the cluster legend of the scatterplot to view that cluster and hide all others. Double click that cluster again in the cluster legend to visualize all clusters again. To view specific details about specific genes, hover over them in the scatterplot

HyPhy aBSREL output per STRING cluster

This table displays the HyPhy aBSREL output data for the A. cahirinus Positively selected genelist for each individual STRING cluster. Select the desired STRING clusters using the range buttons on the left side of the page

This page contains the functional overrepresentation analysis results of both complete aBSREL positively selected genelist and the cluster-specific overrepresentation results The clusterProfiler (4.10.0) R-package was used to perform this functional over-representation analysis. Moreover, we used the rrvgo R package (1.15.1) to reduce the number of cluster-specific significant GO-terms based on their semantic similarity (Sayols, 2023)

Just like the STRING interaction network table, the desired STRING clusters in the table and plot can be selected using the range buttons on the left side of the page. There is also the option to display all the GO-terms per cluster, and not just the reduced GO-terms.



This HyPhy MEME analysis was a follow-up analysis after the HyPhy aBSREL analysis. During the aBSREL analysis, genes were found that exhibited signs of episodic diversifying selection in Acomys cahirinus. However, HyPhy aBSREL detects episodic diversifying selection in specific branches of a phylogeny without informing us which sites are responsible. To get an indication of which codon sites could be under episodic diversifying selection in the genes found by aBSREL, HyPhy MEME was used. HyPhy MEME is a mixed effects model of evolution (MEME) that can detect episodic diversifying selection at an individual site (among all branches). Thus, HyPhy MEME might give us additional insights into which codon sites might be responsible for the episodic diversifying selection signal in the genelist found by aBSREL. A word of caution is advised however, as HyPhy MEME can only detect if a codon site shows signs of diversifying selection among all branches (i.e. MEME detects selection at an individual site, not an individual branch-site). So, HyPhy MEME predicting that a codon site shows significant diversifying selection does not guarantee that any individual branch (like the Acomys cahirinus branch) exhibits diversifying selection at that codon site. nevertheless, HyPhy MEME can give us an indication which sites might be involved in the episodic diversifying selection signal in the Acomys cahirinus genes, even when this limitation is kept in mind.

Every Gene was individually analyzed using HyPhy MEME. The test results are described in the tables below.

This table and summary data represents the MEME likelyhood ratio test results for each site of the selected gene. Here, the positional site information is relative to the human gene that was used in the TOGA multiple sequence alignment (i.e. the multiple sequence alignment with mammalian orthologs of the selected gene, available at the TOGA ortholog database

This tab displays the protein multiple sequence alignment (MSA), which were created by translating the TOGA codon MSAs.

The MSAs have been annotated with the MEME likelyhood ratio test (LRT) P-values. These represent the uncorrected P-values taken straight from the MEME results for the selected gene.

Furthermore, the MSAs are annotated with the A. cahirinus empirical bayes factor (EBF), which gives an indication from which species (i.e. leaf branches in the MEME model) the positive selection signal could be originating from.

Please note that the EBF measure is intended for exploratory purposes only and should not be interpreted as a valid measure of statistical significance.

This page visualizes predicted substitutions at positively selected sites, identified using the HyPhy MEME framework across the mammalian species included in the evolutionary model.

These substitutions are displayed as a tree view for a selected site, or in the form of an interactive table where all substitutions are displayed among the mammalian species The tree view is particularly usefull for identifying where a substitution took place according to the MEME evolutionary model.

Substitutions are in the form of A11Q. Substitutions that appear to be indels are indicated with the term 'Gap' (e.g. C-333-Gap).

Here, the TOGA-Human and TOGA-Acomys reference frames refer to the positions in the TOGA-identified Human and A. cahirinus orthologs respectively. These orthologs can be found at the TOGA ortholog database.

MEME-identified substitutions tree view
All substitutions at positively selected codon sites
Download TSV

For genes with canonical human sequences identified in the UniProtKB/Swiss-Prot database, this page displays MEME-identified positively selected sites mapped onto the canonical isoform.

These Uniprot sequence annotations were retrieved by the EMBL-EBI Proteins REST API, more specifically the features service. The protein visualized below is annotated with the DOMAIN, REGION and MOTIF Uniprot annotation types. Other types of Uniprot annotations can be found in the table below.

Select A. cahirinus empirical Bayes factor (EBF) cutoff value. EBF values below this value will not be displayed in the Uniprot sequence annotation plot and table