Search

Therapeutic peptides and proteins derived from either endogenous hormones, such as insulin, or de novo design via display technologies occupy a distinct pharmaceutical space in between small molecules and large proteins such as antibodies. Optimizing the pharmacokinetic (PK) profile of drug candidates is of high importance when it comes to prioritizing lead candidates, and machine-learning models can provide a relevant tool to accelerate the drug design process. Predicting PK parameters of proteins remains difficult due to the complex factors that influence PK properties; furthermore, the data sets are small compared to the variety of compounds in the protein space. This study describes a novel combination of molecular descriptors for proteins such as insulin analogs, where many contained chemical modifications, e.g., attached small molecules for protraction of the half-life. The underlying data set consisted of 640 structural diverse insulin analogs, of which around half had attached small molecules. Other analogs were conjugated to peptides, amino acid extensions, or fragment crystallizable regions. The PK parameters clearance (CL), half-life (T1/2), and mean residence time (MRT) could be predicted by using classical machine-learning models such as Random Forest (RF) and Artificial Neural Networks (ANN) with root-mean-square errors of CL of 0.60 and 0.68 (log units) and average fold errors of 2.5 and 2.9 for RF and ANN, respectively. Both random and temporal data splittings were employed to evaluate ideal and prospective model performance with the best models, regardless of data splitting, achieving a minimum of 70% of predictions within a twofold error. The tested molecular representations include (1) global physiochemical descriptors combined with descriptors encoding the amino acid composition of the insulin analogs, (2) physiochemical descriptors of the attached small molecule, (3) protein language model (evolutionary scale modeling) embedding of the amino aci

Fast, flexible, and internally valid analytical tools are needed to evaluate the effects of management interventions made on dairy farms to support decisions about which interventions to continue or discontinue. The objective of this observational study was to demonstrate the use of state space models (SSM) to monitor and estimate the effect of interventions on 2 specific outcomes: a dynamic linear model (DLM) evaluating herd-level milk yield and a dynamic generalized linear model evaluating treatment risk in a pragmatic pretest/posttest design under field conditions. This demonstration study is part of a Danish common learning project that ran from March 2020 to May 2021 within the framework of veterinary herd health consultancy in relation to reducing antimicrobial use and improving herd health. Specific interventions for 2 commercial herds were suggested by 4 visiting farmers and were implemented during the project period. The intervention for herd 1 was the application of teat sealers, implemented in August 2020. For herd 2, the intervention was an adjustment of cubicles for cows of parity 2 and above, implemented from November 2020. A shift to an automatic milking system in October 2020 was also modeled as an intervention for herd 1 because the 2 interventions coincided. Data available from the Danish Cattle Database on obligatory registrations for individual cow movements and treatments, as well as test day information on milk yield, were used for model building and testing. Data from a 3-yr period before the project were used to calibrate the SSM to herd conditions, and data from the study period (March 2020 to May 2021) were used for monitoring and intervention testing based on application of the SSM. Herd bulk tank milk recordings were added to the data set during the study period to increase the precision of the estimates in the DLM. The developed SSM monitored herd-level milk yield and the overall probability of treatment throughout the study period in