Title Evaluation of genome-based growth predictions in Microbacterium
Advisor Dr. Luis M. Rodriguez-R, Department of Microbiology and Digital Science Center (DiSC)
Co-advisor Dr. Sigrid Neuhauser, Department of Microbiology
Number of students 1
Language English


It has been previously shown that both the maximum growth rate and the optimal growth temperature of bacteria imprints statistical signatures in their genomes. Specifically, the codon bias of highly-transcribed genes such as ribosomal genes with respect to the average codon usage of the rest of the genome appears to predict the minimum doubling time of bacteria. We have generated a large collection of growth curves from isolates of different species in the genus Microbacterium, using controlled, identical conditions, and have determined their corresponding genome sequences. The selected student will have access to this data, and will evaluate the fit of growth curves to determine both the minimum doubling time and the range of incertitude. Next, the student will generate growth predictions on the basis of codon usage biases with existing statistical models, and will provide a systematic evaluation of their performance in this collection. Finally, the student will be able to explore the functional annotation and other genomic signatures in order to identify additional sources of information that could explain the prediction residuals, potentially improving the growth rate predictions in other taxa.


Ongoing Master’s degree in microbiology, a capacity for both independent and team work, and interest in both molecular biology and genomics.

Theoretical skills

Foundations on microbiology and molecular biology are highly desirable. Interest in environmental genomics is required, but no previous knowledge necessary. Strong foundations in statistics are highly appreciated.

Practical skills

Basic knowledge on R and Bash are preferred, but not required.