Validation of novel Parkinson's Disease target genes by Artificial Intelligence-based predictions

Parkinson's disease (PD) is the second most common age related disorder and the fastest growing
neurological condition in the world, with no cure. PD is characterised by a progressive loss of dopaminergic
neurons in the substantia nigra pars compacta, leading to impaired movement such as tremors, stiffness
and slowness of movements. There is an urgent need for new treatment options that address the underlying
disease mechanisms and improve upon the current therapeutic approach of managing symptoms. Recent
advances in our understanding of the causes of PD have highlighted a central role for mitochondrial and
lysosomal dysfunction in the development of the disease (Wallings et al, 2019). The aim of this PhD
studentship is to identify and validate genes involved in PD progression derived from hypothesis generation
by Artificial Intelligence (AI) in the context of lysosomal and mitochondrial biology. As post-mitotic cells
appropriate, lysosomal and mitochondrial function is vital to neuron viability because of their involvement in
recycling of waste material and energy production. PD, like other neurodegenerative diseases, is
characterised by an accumulation of protein aggregates highlighting deficiencies in the clearance pathways.
Genetic studies have established an important link between lysosomal and mitochondrial dysfunction and
the pathogenesis of PD. A large number of autosomal dominant and recessive genes are associated with
PD as well as several genetic risk factors which encode for key proteins involved in mitochondrial quality
control and lysosomal activity, including PINK1 or GBA1. In healthy cells, damaged mitochondria are
removed from cells by mitophagy, i.e., the capture of damaged mitochondria by autophagosomes with
subsequent delivery to the lysosome. This process and successful degradation of damaged mitochondria by
the lysosome is defective in many forms of PD. The main aim of this project is to identify genes that can be
manipulated to enhance mitophagy and lysosomal function and thus improve dopaminergic neuron survival.
The successful student will be supported by a multidisciplinary team, combining the strengths in functional
genomics and imaging-based screening (Ketteler lab) with target identification from AI and validation in
neurodegenerative diseases (BenevolentAI). The project will proceed in two distinct phases:
First, AI-driven identification of targets by BenevolentAI (this work will be completed before the studentship
start date), followed by siRNA- and CRISPR/Cas9-based screening of these targets in functional mitophagy
assays (at UCL). Secondly, the validation of identified hits from the mechanistic screening in
disease-relevant neuronal cells derived from induced pluripotent stem cells (developed with BenevolentAI
and UCL). We anticipate that 12 months of this work will take place in research labs at BenevolentAI, thus
enabling the student to gain significant experience in working in an industry environment.
During this interdisciplinary project the student will acquire skills in quantitative biology, functional genomics,
cell biology and early stage drug discovery and will be able to get insight into current approaches in AI and
machine learning. We believe that the computational prediction of novel targets by BenevolentAI combined
with Ketteler's expertise in systematic analysis of mitophagy and neurodegeneration places us in a unique
position to identify novel aspects of PD pathogenesis.