Machine learning methods to study sequence-ensemble-function relationships in disordered proteins

The thermodynamic hypothesis of protein aggregation

Protein misfolding and aggregation drive some of the most prevalent and lethal disorders of our time, including Alzheimer's and Parkinson's diseases, now affecting tens of millions of people worldwide. The complexity of these diseases, which are often multifactorial and related to age and lifestyle, has made it challenging to identify the causes of the accumulation of aberrant protein deposits. An insight into the origins of these deposits comes from reports of a widespread presence of protein aggregates even under normal cellular conditions. This observation is best accounted for by the thermodynamic hypothesis of protein aggregation. According to this hypothesis, many proteins are expressed at levels close to their supersaturation limits, so that their native states are metastable against aggregation. Here we integrate the evidence behind this hypothesis and outline actionable therapeutic strategies that could halt protein aggregation at its source.