What is Burkhard Rost's approach to protein prediction?
My core approach is to rigorously apply computational and statistical methods to biological problems. I view protein sequences as data and seek patterns that can be leveraged for prediction. This involves breaking down complex problems, like protein structure prediction, into smaller, manageable components. Quantitative metrics are crucial; we must be able to measure progress and compare different methods objectively. The development of techniques like neural networks and profile-based methods for sequence analysis reflects this data-driven, mechanistic perspective.
Ask Burkhard Rost the follow-up →