Yesterday eLife published the first paper from our citizen science project, BashTheBug, which was launched in April 2017 on the Zooniverse platform. Through BashTheBug we asked for volunteers to classify images of M. tuberculosis growing on a range of concentrations of 13 different antibiotics. The images were derived from photographs of 96-well plates taken after […]
Category: tuberculosis
BashTheBug is a citizen science project hosted on the Zooniverse platform that we launched in April 2017 and asked volunteers to help us assess how well 20,637 clinical samples of M. tuberculosis grow on one of 13 different antibiotics. To help engage with the volunteers it has its own blog, that has grown into the […]
The CRyPTIC project collecting over 20,000 clinical samples of TB and for each, sequencing its genome and testing its susceptibility to 13 different antibiotics. A lovely unintended consequence of compiling such a large high-quality dataset is that CRyPTIC was invited to form part of the team that collected data and compiled the first catalogue of […]

In this preprint, which Alice has been working on for several years, we show how alchemical free energy methods can predict whether an amino acid mutation confers resistance to an antitubercular, but only in cases where the change in binding free energy is large. This is mainly because the confidence limits on the change in […]

Oliver Adams successfully elucidated the structure of the M. tuberculosis MmpL3 membrane transporter using cryo-EM and this has recently been published online in Structure. This was the main aim of his PhD studies in Simon Newstead‘s group in the Department of Biochemistry here in Oxford. It is an important protein structure since although other MmpL3 […]

Through the CompBioMed2 EU Centre of Excellence project I have funding to appoint a postdoctoral researcher to develop machine-learning models to predict whether an infection is susceptible to an antibiotic. The need for predictive methods, such as these, will grow in the coming years as more of clinical microbiology transitions to using genetics to infer […]

In this preprint, the CRyPTIC project proposes the maximum value of minimum inhibitory concentration (MIC) for 13 different anti-TB drugs below which a sample can be considered to be ‘genotypically wild-type’. It is necessary to establish these values, called epidemiological cutoff values (ECOFFs or ECVs), so that the MICs measured can be converted into binary […]

Although the population structure M. tuberculosis is clonal, one must be careful when inferring the effect of individual mutations on the effect of an antibiotic. Purely because a mutation appears to define a phylogeny does not mean it has no effect on the minimum inhibitory concentration. Read more here (Open Access).

AMyGDA is a python module that analyses photographs of 96-well plates and, by examining each well for bacterial growth, is able to read a series of minimum inhibitory concentrations for the antibiotics present on a plate. Previously it was only available to download from this website (due to licensing) if you gave your email address […]
Numpy v Biopython

Having only recently having to write bioinformatics Python code that e.g. interrogate GenBank files to find out the sequence of specific genes I’ve learnt a bit of Biopython. I’ve always wondered why (and I could be wrong) the bioinformatics community doesn’t make more use of numpy? For example the Seq class in Biopython seems to be […]