Dr David Martin

David Martin's research focused on high throughput analysis of large data sets. One major challenge of modern biology is dealing with the increasing rate at which data is produced, and extracting biological meaning from it. This data needs to be captured, managed, interpreted and represented in ways which allow inferences to be drawn by experimentalists.

David spends most of his time teaching but former projects include the following:

Automated interpretation of post-translational modification by MS-MS

Protein phosphorylation (the post-translational modification of a protein by adding a phosphate ion at a specific point in a polypeptide chain) is a key mechanism that is used by the cell to control processes such as growth and cell death. In collaboration with Professor Mike Ferguson, we are performing a global phosphoproteome analysis of the Trypanosoma brucei parasite.

Mass spectrometry can identify proteins through accurate measurement of the mass of peptide fragments. Subsequent fragmentation of specific peptide ions provides a tandem MS (MS-MS) spectrum, allowing the peptide sequence to be determined following database searching with appropriate software such as MASCOT. This software provides a list of matches in the database and highlights the putative position of any post-translational modification. In order to verify the database match, the MS-MS spectrum must be inspected by an expert as the database search algorithms are not designed for post-translational modification identification. However, complex protein mixtures such as whole-cell lysates can provide many hundreds of thousands of spectra, manual validation of which is extremely time-consuming.

To automate this process, a post-search validation method has been developed. The database match is re-evaluated against a set of criteria which model the expected chemical behaviour of the peptide fragments. The criteria are stringent so the method provides high confidence hits, allowing the expert to assess the harder spectra which are at present unable to be assessed manually. With this methodology, we can at present automatically identify approximately half the phosphorylation sites with negligible false positives. There is plenty of potential for further improvement, giving the possibility of rapid, high throughput phosphoproteome scans.

Sequencing Chromosome 4 of the Potato genome

Genome sequencing raises considerable genomic challenges. Integration of genetic and physical map data, mapping of clones and assembly of sequence data require considerable bioinformatics support. In addition, next-generation sequencing technologies such as Solexa and 454 have the potential to radically change the approach taken to obtaining novel genomes.

In collaboration with colleagues at the Scottish Crop Research Institute (Dr Glenn Bryan), Teagasc Ireland (Dr Dan Milbourne) and Imperial College London (Dr Gerard Bishop), we will be determining the genomic sequence of potato (Solanum tuberosum) chromosome 4.

The project provides many bioinformatic challenges and opportunities. The strain being sequenced is heterozygous, adding complexity to the construction of tiling paths and data integration. There are many strains with different traits which can be mapped to regions of interest, and Potato is closely related to Tomato allowing for direct interspecific comparison. Data at present has been determined using classical Sanger sequencing in a BAC by BAC strategy. We will be exploring the utility of NGS data, both for BAC by BAC approaches and whole-genome shotgun assemblies. In addition, new methods and techniques are being developed.

This project is funded by BBSRC grant  BB/F012640/1.

GOtcha - providing qualified functional assignments.

GOtcha [4] is a methodology for rapid functional annotation of gene products. Gene Ontology is a hierarchical description of the function of a gene product. Specific terms are linked to more general terms, providing a tree-like structure. GOtcha makes use of this hierarchy to examine the functional assignment of similar gene products and provide a likelihood for the sequence in question to have that particular function.

GOtcha assignments for an entire genome can be mapped to metabolic pathways, indicating biological processes which may be present or absent in that organism. GOtcha was used in the analysis of the genome data for Plasmodium falciparum (Malaria), Trypanosoma brucei (African Sleeping Sickness), Trypanosoma cruzi (Chagas disease), Leishmania major (Leishmaniasis), and Brugia malayi (elephantiasis).

• GOtcha Search

Kinomer- Detailed classification of protein kinases

In collaboration with Dr Diego Miranda-Saavedra (now at Cambridge University). Application of subgroup-specific Hidden Markov Models for classification and identification of protein kinases.

• Kinomer

Detailed classification of SNF2-like helicases

In collaboration with Dr Andrew Flaus (now at University of Galway, Ireland). Calculation and application of subgroup-specific Hidden Markov Models for classification and identification of SNF2-like [1] chromatin remodelers.

David is part of the core level 1 & 2 teaching team, delivering a core curriculum across all Life Sciences degrees. His particular focus is on numeracy and data literacy, bringing quantitative methodologies into the approach that students take as part of their developing roles as active scientists. Students are encouraged to build up their data analysis skills as a key tool in enabling them to question and evaluate the world around them.

Statistics/Data analysis:

We use Rstudio as the core technology for all our data analysis and statistical analysis. Reproducibility is key to effective science so students are encouraged to include their R scripts in project reports and analyses so that errors or misunderstandings can be cleared up, or so we can learn new ways of doing things when they have found something really cool.

Techniques and Tools

Wherever possible I like students to perform real research instead of 'toy' practicals. We should contribute to knowledge as we learn.

We use Jalview extensively as a workbench for sequence analysis and comparison form the beginning of our course. In addition, the chimaera is used for protein visualisation. Where possible, all the data analysis and visualisation software are cross-platform and freely available - tools and knowledge the students can take with them.

I introduce students to a taster of Python programming in year 2. For those keen to pursue this further, the bioinformatics 3rd-year modules extend skills and abilities in Python and R  applied to current activities including next-generation sequencing to analyse gene expression. In the fourth year, we broaden knowledge by looking at the application of bioinformatics algorithms across different subject areas.
 

Student Projects:

I supervise a number of students in their Honours year project. This can be a  lab project where we apply bioinformatics to a real problem and maybe test our hypotheses in the lab or science communication where students develop and deliver educational activities to an appropriate audience.

Selected recent student projects: 

• Identification and validation of novel microsatellite markers in the Soprano Pipistrelle Pipistrellus pygmaeus
• A card game to educate on crop development and GMO technology

Key Teaching Achievements:

Nominated for Student-Led Teaching Awards 2014

Produced 'Lost in Translation', a board game for 1st year/A2 students to reinforce understanding of DNA transcription and translation

Produced Gigsaw - a teaching tool for understanding Next Generation Sequence analysis