Siems, K. | bioRxiv Science

bioRxiv2026

Site-Specific Introduction of Non-Canonical Amino Acids into natural and engineered Non-Ribosomal Peptides

The incorporation of non-canonical amino acids (ncAAs) into proteins, developed in the past 20 years, has opened new avenues with respect to protein structure, protein modification, protein-protein interaction or enzyme catalysis beyond what is possible with the 20 proteinogenic AAs. Although >300 unusual building blocks including several ncAAs have been described in nonribosomal peptides (NRPs) naturally, we aimed to further expand the scope of the underlying nonribosomal peptide synthetases (NRPS) to incorporate ncAAs beyond the naturally available ones. We have therefore systematically screened for ncAA accepting NRPS systems, applied NRPS engineering to transfer the respective ncAA-accepting parts into other NRPSs and thereby created novel peptides that were further derivatized in post-enzymatic chemical synthesis reactions directly in bacterial culture extracts. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=177 SRC="FIGDIR/small/738027v1_ufig1.gif" ALT="Figure 1"> View larger version (38K): org.highwire.dtl.DTLVardef@8968ddorg.highwire.dtl.DTLVardef@143a852org.highwire.dtl.DTLVardef@dfac73org.highwire.dtl.DTLVardef@b2c6fb_HPS_FORMAT_FIGEXP M_FIG C_FIG

biochemistry↗

bioRxiv2020

Intense bitterness of molecules: machine learning for expediting drug discovery

Drug development is a long, expensive and multistage process geared to achieving safe drugs with high efficacy. A crucial prerequisite for completing the medication regimen for oral drugs, particularly for pediatric and geriatric populations, is achieving taste that does not hinder compliance. Currently, the aversive taste of drugs is tested in late stages of clinical trials. This can result in the need to reformulate, potentially resulting in the use of more animals for additional toxicity trials, increased financial costs and a delay in release to the market. Here we present BitterIntense, a machine learning tool that classifies molecules into "very bitter" or "not very bitter", based on their chemical structure. The model, trained on chemically diverse compounds, has above 80% accuracy on several test sets. BitterIntense suggests that intense bitterness does not correlate with toxicity and hepatotoxicity of drugs and that the prevalence of very bitter compounds among drugs is lower than among microbial compounds. BitterIntense allows quick and easy prediction of strong bitterness of compounds of interest for food and pharma industries. We estimate that implementation of BitterIntense or similar tools early in drug discovery and development process may lead to reduction in delays, in animal use and in overall financial burden. Significance StatementDrug development integrates increasingly sophisticated technologies, but extreme bitterness of drugs remains a poorly addressed cause of medicine regimen incompletion. Reformulating the drug can result in delays in the development of a potential medicine, increasing the lead time to the patients. It might also require the use of extra animals in toxicity trials and lead to increased costs for pharma companies. We have developed a computational predictor for intense bitterness, that has above 80% accuracy. Applying the classifier to annotated datasets suggests that intense bitterness does not correlate with toxicity and hepatotoxicity of drugs. BitterIntense can be used in the early stages of drug development to identify drug candidates that require bitterness masking, and thus reduce animal use, time and monetary loss.

pharmacology and toxicology↗