sanjaynagi / anoexpress Goto Github PK

potentially allowing 'gamb', colu, arab as analysis options.

otherwise add optional taxon command

• We should allow users to query based on annotations (pfam, go) within the data() function. This will simplify code in other functions too.

• Fix spelling of Gaura
• change boauke to half coluzzii/gambiae colours

• Figure 1 - Update / correct sample map

• Figure 2 - GWES (matplotlib/seaborn)

• Figure 3 - Heatmap of highest median log2 FC (up)

• Figure 4 - Heatmap of most consistent genes (up)

• Figure 5 - Heatmaps for detox genes A) P450s, B) other detox C) other

• Figure 6 - Dot plot venn of genes

• Suppl Figure 1 - PCA

I didnt know that Colab supports R.

This means if we move the differential expression analysis into a Colab-able version, the ENTIRE AnoExpress analysis can be done from Colab, linked to from the README, which is very cool and extreme levels of reproducibility.

• Full dataset (~13000 genes), not reduced due to finding orthologs (8599 genes) in funestus

• requires using plotly subplots

• Find genes with high median FCs etc across experiments
• Find genes consistently up/down regulated

• Currently fcs, pvals have GeneID and GeneName
• but counts only has GeneID

Which makes it harder to handle both data types with a single line of code

Plot gene expression for a whole contig, which can be filtered by p-value.

Great for quickly scanning genomic regions for evidence of IR genes/mutations

with pytest

May change the overall results a lot, but could be worth it. Shrinks FCs for very lowly expressed genes, which is kind of an ideal behaviour.

Will not implement until Version 2.

as an option to include old microarray studies.

https://www.nature.com/articles/s41467-018-07615-x from this study.

I already should have the data, just need to make sure we have metadata.

as well as the current options - agap and median

now we need to have authorisation to use malariagen_data, it makes things complicated for users of anoexpress.

We currently use ag3 for two things, to load the gff file, and to plot the genes track for the GWES analysis.

For now, lets load the gff from vectorbase and use that.

• Check for specific values in IRTex loadings
• and other data
• top candidates

One for comparisons and one for samples

Users could provide a region string - '2R:28500000-28600000' to the geneID parameter to plot expression for all genes within that region.

• first need countries/locations of each experiment

• Count number of orthologs per gene
• Effect on expression

• do we remove tiassale as it will not be published soon