Alix, J. | Search bioRxiv

bioRxiv2024

predicTTE: An accessible and optimal tool for time-to-event prediction in neurological diseases

Time-to-event prediction is a key task for biological discovery, experimental medicine, and clinical care. This is particularly true for neurological diseases where development of reliable biomarkers is often limited by difficulty visualising and sampling relevant cell and molecular pathobiology. To date, much work has relied on Cox regression because of ease-of-use, despite evidence that this model includes incorrect assumptions. We have implemented a set of deep learning and spline models for time-to-event modelling within a fully customizable app and accompanying online portal, both of which can be used for any time-to-event analysis in any disease by a non-expert user. Our online portal includes capacity for end-users including patients, Neurology clinicians, and researchers, to access and perform predictions using a trained model, and to contribute new data for model improvement, all within a data-secure environment. We demonstrate a pipeline for use of our app with three use-cases including imputation of missing data, hyperparameter tuning, model training and independent validation. We show that predictions are optimal for use in downstream applications such as genetic discovery, biomarker interpretation, and personalised choice of medication. We demonstrate the efficiency of an ensemble configuration, including focused training of a deep learning model. We have optimised a pipeline for imputation of missing data in combination with time-to-event prediction models. Overall, we provide a powerful and accessible tool to develop, access and share time-to-event prediction models; all software and tutorials are available at www.predictte.org.

neuroscience↗

bioRxiv2024

Maximising the translational potential of neurophysiology in amyotrophic lateral sclerosis: a study on compound muscle action potentials

Transgenic mouse models of amyotrophic lateral sclerosis, such as the widely used SOD1G93A mouse, enable investigation of disease mechanisms and testing of novel therapeutic interventions. However, treatments that have been considered successful in mice have often failed to translate into human benefit in clinical trials, particularly when relying on the so-called survival read-out. Compound muscle action potentials (CMAPs), are a simple neurophysiological test that measures the summation of muscle fibre depolarisation in response to maximal stimulation of the innervating nerve. CMAPs can be measured in both mice and humans and decline with motor axon loss in ALS, making them a potential translational read-out of disease progression which could help bridge the preclinical and clinical divide. Herein we assess the translational potential of CMAPs and ascertain at what time points human and mouse data aligned most closely. We extracted data from 18 human studies and compared with results generated from SOD1G93A and control mice at different ages across different muscles. We found that the relative CMAP amplitude difference between SOD1G93A and control mice in tibialis anterior and gastrocnemius muscles at 70 days of age was most similar to the relative difference between baseline ALS patient CMAP measurements and healthy controls in the abductor pollicis brevis (APB) muscle. We also found that the relative decline in SOD1G93A tibialis anterior CMAP amplitude between 70-140 days was similar to that observed in 12 month human longitudinal studies in APB. Our findings suggest CMAP amplitudes can provide a translational window, from which to make comparisons between the SOD1G93A model and human ALS patients. CMAPs are easy to perform and can help determine the most clinically relevant starting/end points for preclinical studies and provide a basis for predicting potential clinical effect sizes.

neuroscience↗