This week we’ll be discussing potential project ideas that we might like to tackle as a group. You are encouraged to bring along your ideas, but to get the discussion started James, Rob and I (Darcy) have come up with a few ideas for projects that we think might be appropriate.

Codon adaptation indices

Deviations from normal codon usage may indicate some compartmentalisation of mutational load (e.g. RIP) or presence of foreign DNA (e.g. horizontal transfer). Last year James had a volunteer (Renu) develop a method for finding codon adaptation indices (CAI). This project would involve packaging that method into some software suitable for distribution, and would probably also involve a web interface.

This is a fairly simple project so is probably a good choice for an initial project, but may be too small for the number of people that we have.

Updating RIPCAL

RIPCAL is a tool written by James for alignment bases analyses of RIP in genomes. The tool is still very useful but was developed more than a decade ago, before the rise of short read sequencing. Because repeated sequences are particularly difficult to assemble using Illumina sequencing technologies, there is a need for an assembly agnostic method of RIP characterisation. Although long read sequencing is returning, resequencing efforts and pangenomics are likely to rely on the cheaper options for some time to come.

So there are two options for updating RIPCAL.

1) Update the software to target newer generation high quality genomes, using updated tools (e.g. ClustalW -> ClustalO), include unittests, and have a web interface. 2) Develop a new method to target fragmented genomes, which will likely take either an alignment free (i.e. kmer counting/clustering followed by local assembly), or a read alignment (BAM/SAM) approach.

Both tools will be useful and publishable. Both will require a diverse set of tasks, including some basic web development. The second option is obviously more complex, which might be more or less appealing to you.

Web based genomic visualisations

In fungal genomics we often need customised plots to investigate genomic rearrangements (e.g. mesosynteny). Unfortunately rich customisable genomic plots are often difficult to produce, even with the powerful plotting libraries available. This project would build involve developing pipelines to plot complex (possibly interactive) visualisations using web technologies. Rob has previously presented some genome alignments using Javascript, D3, and other web standards. We would be building on that experience.

Depending on the compenteny, interest level, and optimism of participants, there might be an opportunity to improve performance of interactive components using the new Web assembly standard.

Each of these projects should have aspects suited to beginner and experienced programmers, and a diverse enough set of tasks to allow most of us to fill in knowledge gaps. The projects all also include some level of web development which is becoming more and more relevant to engaging non-specialists in research software.

Again, please bring along your own project ideas, especially if they address a specific research question that you have but might be too big for you to tackle alone. Our examples are pretty bioinformatics focussed, but yours don’t have to be.

Cheers, Darcy.