MolPLA: a molecular pretraining framework for learning cores, R-groups and their linker joints.

MOTIVATION: Molecular core structures and R-groups are essential concepts in drug development. Integration of these concepts with conventional graph pre-training approaches can promote deeper understanding in molecules. We propose MolPLA, a novel pre-training framework that employs masked graph contrastive learning in understanding the underlying decomposable parts in molecules that implicate their core structure and peripheral R-groups. Furthermore, we formulate an additional framework that grants MolPLA the ability to help chemists find replaceable R-groups in lead optimization scenarios. RESULTS: Experimental results on molecular property prediction show that MolPLA exhibits predictability comparable to current state-of-the-art models. Qualitative analysis implicate that MolPLA is capable of distinguishing core and R-group sub-structures, identifying decomposable regions in molecules and contributing to lead optimization scenarios by rationally suggesting R-group replacements given various query core templates. AVAILABILITY AND IMPLEMENTATION: The code implementation for MolPLA and its pre-trained model checkpoint is available at https://github.com/dmis-lab/MolPLA.

ArkDTA: attention regularization guided by non-covalent interactions for explainable drug-target binding affinity prediction.

MOTIVATION: Protein-ligand binding affinity prediction is a central task in drug design and development. Cross-modal attention mechanism has recently become a core component of many deep learning models due to its potential to improve model explainability. Non-covalent interactions (NCIs), one of the most critical domain knowledge in binding affinity prediction task, should be incorporated into protein-ligand attention mechanism for more explainable deep drug-target interaction models. We propose ArkDTA, a novel deep neural architecture for explainable binding affinity prediction guided by NCIs. RESULTS: Experimental results show that ArkDTA achieves predictive performance comparable to current state-of-the-art models while significantly improving model explainability. Qualitative investigation into our novel attention mechanism reveals that ArkDTA can identify potential regions for NCIs between candidate drug compounds and target proteins, as well as guiding internal operations of the model in a more interpretable and domain-aware manner. AVAILABILITY: ArkDTA is available at https://github.com/dmis-lab/ArkDTA. CONTACT: kangj@korea.ac.kr.

Metabolic engineering of Escherichia coli for production of mixed isoprenoid alcohols and their derivatives.

BACKGROUND: Current petroleum-derived fuels such as gasoline (C5-C12) and diesel (C15-C22) are complex mixtures of hydrocarbons with different chain lengths and chemical structures. Isoprenoids are hydrocarbon-based compounds with different carbon chain lengths and diverse chemical structures, similar to petroleum. Thus, isoprenoid alcohols such as isopentenol (C5), geraniol (C10), and farnesol (C15) have been considered to be ideal biofuel candidates. NudB, a native phosphatase of Escherichia coli, is reported to dephosphorylate isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP) into isopentenol. However, no attention has been paid to its promiscuous activity toward longer chain length (C10-C15) prenyl diphosphates. RESULTS: In this study, the promiscuous activity of NudB toward geranyl diphosphate (GPP) and farnesyl diphosphate (FPP) was applied for the production of isoprenoid alcohol mixtures, including isopentenol, geraniol, and farnesol, and their derivatives. E. coli was engineered to produce a mixture of C5 and C15 alcohols by overexpressing NudB (dihydroneopterin triphosphate diphosphohydrolase) and IspA (FPP synthase) along with a heterologous MVA pathway, which resulted in a total of up to 1652 mg/L mixture of C5 and C15 alcohols and their derivatives. The production was further increased to 2027 mg/L by overexpression of another endogenous phosphatase, AphA, in addition to NudB. Production of DMAPP- and FPP-derived alcohols and their derivatives was significantly increased with an increase in the gene dosage of idi, encoding IPP isomerase (IDI), indicating a potential modulation of the composition of the alcohols mixture according to the expression level of IDI. When IspA was replaced with its mutant IspA*, generating GPP in the production strain, a total of 1418 mg/L of the isoprenoid mixture was obtained containing C10 alcohols as a main component. CONCLUSIONS: The promiscuous activity of NudB was newly identified and successfully used for production of isoprenoid-based alcohol mixtures, which are suitable as next-generation biofuels or commodity chemicals. This is the first successful report on high-titer production of an isoprenoid alcohol-based mixture. The engineering approaches can provide a valuable platform for production of other isoprenoid mixtures via a proportional modulation of IPP, DMAPP, GPP, and FPP syntheses.

Isoprene production by Escherichia coli through the exogenous mevalonate pathway with reduced formation of fermentation byproducts.

BACKGROUND: Isoprene, a volatile C5 hydrocarbon, is an important platform chemical used in the manufacturing of synthetic rubber for tires and various other applications, such as elastomers and adhesives. RESULTS: In this study, Escherichia coli MG1655 harboring Populus trichocarpa isoprene synthase (PtispS) and the exogenous mevalonate (MVA) pathway produced 80 mg/L isoprene. Codon optimization and optimal expression of the ispS gene via adjustment of the RBS strength and inducer concentration increased isoprene production to 199 and 337 mg/L, respectively. To augment expression of MVA pathway genes, the MVA pathway was cloned on a high-copy plasmid (pBR322 origin) with a strong promoter (Ptrc), which resulted in an additional increase in isoprene production up to 956 mg/L. To reduce the formation of byproducts derived from acetyl-CoA (an initial substrate of the MVA pathway), nine relevant genes were deleted to generate the E. coli AceCo strain (E. coli MG1655 ΔackA-pta, poxB, ldhA, dld, adhE, pps, and atoDA). The AceCo strain harboring the ispS gene and MVA pathway showed enhanced isoprene production of 1832 mg/L in flask culture with reduced accumulation of byproducts. CONCLUSIONS: We achieved a 23-fold increase in isoprene production by codon optimization of PtispS, augmentation of the MVA pathway, and deletion of genes involved in byproduct formation.

Farnesol production in Escherichia coli through the construction of a farnesol biosynthesis pathway - application of PgpB and YbjG phosphatases.

Farnesol is a sesquiterpenoid alcohol that has important industrial and medical potential. It is usually synthesized from farnesyl diphosphate (FPP) by farnesol synthase in plants. FPP accumulation can cause up-regulation of phosphatases capable of FPP hydrolysis, resulting in farnesol production in Escherichia coli. We found that PgpB and YbjG, two integral membrane phosphatases, can hydrolyze FPP into farnesol. Overexpression of FPP synthase (IspA) and PgpB, along with a heterologous mevalonate pathway, enabled recombinant E. coli to produce 526.1 mg/L of farnesol. This result indicates that the phosphatases PgpB and YbjG can be used to construct a novel farnesol synthesis pathway for mass production in E. coli.