My long-term goal is to contribute to a deeper understanding of protein's functions, including their evolution, folding, interactions and associated diseases in humans. Key to my success in this area has been the consideration and integration of diverse sources of information, servers and databases, relevant to predicted protein functions and interactions. Trying to build consistent bridges between protein sequence, structure, and molecular function allows me to establish reliable links between genotype and phenotype. My most important scientific achievements:
- Description of new protein domains, such as: Brichos, Potra, Marvel, Rawul, Acrata, or Spoc.
- Identification of remote orthologous relationships (for example: Treslin/Sld3 and HJURP/Scm3)
- Prediction of novel enzymatic activities in diverse families of proteins of unknown function (for example: FTO, Cdc45, Vasohibin, TMEM6SF2/MAC30, or TIKI families).
Independently, my current interests range from (I) the evolutionary study of transcriptional regulatory complexes in eukaryotes to (II) the analysis of genes and pathways associated with human diseases.
- Predicting and determining domain functions in transcription (MRC Core programme). Transcription initiation and elongation in mammals is a complex process involving dozens of molecules. Despite the importance of understanding this process it remains unclear what proteins are essential for specific molecular activities, and how these relate evolutionarily to proteins studied in model organisms. In this project we are using cutting-edge sequence comparison approaches to predict the functions and evolutionary heritage of domains within transcription-related molecules, focusing initially on one with likely nucleic acid-binding function. Subsequently, computationally-derived hypotheses are being addressed experimentally in order to determine molecular activities.
- Genes and pathways associated with human diseases (Mendelian and non-Mendelian). The discovery of sequence and structural similarities among proteins is a powerful approach to infer functional and evolutionary relationships for known and unknown proteins. This can lead to experimentally-tractable hypotheses and result in a better understanding of protein functions under non-pathological and pathological conditions.