Wisdom of crowds for synthetic accessibility evaluation.

Synthetic accessibility evaluation is a process to assess the ease of synthesis of compounds. A rapid method for the assessment of synthetic accessibility for a vast number of chemical compounds is expected to bring about a breakthrough in the drug discovery. Although several computational methods have been proposed, the compound evaluation has still been processed by medicinal chemists; however, the low throughput of the human evaluation due to the lack of chemists is a critical issue for handling a large number of compounds. We propose the use of crowdsourcing for addressing this problem, and we conducted experiments to investigate the feasibility of incorporating semi-experts and a statistical aggregation method into the synthetic accessibility evaluation. Our experimental results show that we can obtain accurate synthetic accessibility scores through the statistical aggregation of judgments from semi-experts.

Bayesian molecular design with a chemical language model.

The aim of computational molecular design is the identification of promising hypothetical molecules with a predefined set of desired properties. We address the issue of accelerating the material discovery with state-of-the-art machine learning techniques. The method involves two different types of prediction; the forward and backward predictions. The objective of the forward prediction is to create a set of machine learning models on various properties of a given molecule. Inverting the trained forward models through Bayes' law, we derive a posterior distribution for the backward prediction, which is conditioned by a desired property requirement. Exploring high-probability regions of the posterior with a sequential Monte Carlo technique, molecules that exhibit the desired properties can computationally be created. One major difficulty in the computational creation of molecules is the exclusion of the occurrence of chemically unfavorable structures. To circumvent this issue, we derive a chemical language model that acquires commonly occurring patterns of chemical fragments through natural language processing of ASCII strings of existing compounds, which follow the SMILES chemical language notation. In the backward prediction, the trained language model is used to refine chemical strings such that the properties of the resulting structures fall within the desired property region while chemically unfavorable structures are successfully removed. The present method is demonstrated through the design of small organic molecules with the property requirements on HOMO-LUMO gap and internal energy. The R package iqspr is available at the CRAN repository.

Significance of plasma adiponectin for diagnosis, neurological severity and functional outcome in ischemic stroke - Research for Biomarkers in Ischemic Stroke (REBIOS).

OBJECTIVE: Although adiponectin is a major adipocytokine that affects the pathogenesis of various cardiovascular diseases, its clinical significance in stroke remains controversial. We investigated the clinical significance of plasma adiponectin for the diagnosis, neurological severity and functional outcomes of patients with ischemic stroke. METHODS: We prospectively enrolled 171 patients with ischemic stroke and 171 age- and sex-matched healthy controls. Blood samples and clinical information were obtained at day 0, 3, 7, 14 and 90 after stroke onset. RESULTS: Average adiponectin values at day 0 did not significantly differ between the controls and the patients, but were significantly lower and higher in patients with atherothrombotic brain (ATBI) (p=0.047) and cardioembolic (CE) (p=0.008) infarction, respectively, than in the controls. Multivariate logistic regression analyses showed that the adiponectin value at day 0 could predict ATBI (odds ratio, 0.75; 95% confidence interval, 0.58 to 0.91, p=0.009, per 1-µg/mL increase). Adiponectin values at day 0 were positively associated with neurological severity as evaluated by the National Institute of Health Stroke Scale upon admission (r=0.420, p=0.003) and were higher in the groups with poor outcomes (modified Rankin Scale (mRS) ≥ 3 on day 90) than in those with good ones (mRS ≤ 2) in all stroke subtypes, with statistical significance in ATBI (p=0.015). CONCLUSIONS: Plasma adiponectin values may help to classify stroke subtypes and predict neurological severity and functional outcome in ischemic stroke patients.

Plasma S100A12 is associated with functional outcome after ischemic stroke: Research for Biomarkers in Ischemic Stroke.

BACKGROUND: Ischemic stroke is accompanied by an inflammatory response, which exacerbates brain injury and deteriorates functional outcome. S100A12 is expressed abundantly in granulocytes, and has been implicated to play an important role on inflammatory reactions in various disease states. We aimed to determine the association between plasma S100A12 levels and a functional outcome in patients with acute ischemic stroke. METHODS: We prospectively included 171 patients with acute ischemic stroke within 24h after onset in this study. Plasma samples were collected for the measurement of S100A12 levels. Poor functional outcome was defined as a modified Rankin Scale of 2-6 at day 90 after stroke onset. RESULTS: Of 171 patients, 74 (43.3%) had a poor functional outcome at day 90 after stroke onset. Plasma S100A12 levels on admission were significantly higher in patients with a poor functional outcome (2.1 [1.2-5.1] ng/mL, median [interquartile]) than in those with a favorable outcome (1.1 [0.5-2.0] ng/mL; p<0.001). Multivariate analysis showed that the highest quartile of plasma S100A12 levels on admission showed a significantly higher risk for a poor functional outcome (odds ratio, 4.01; 95% confidence interval, 1.09-16.10; p=0.03) than the lowest quartile. CONCLUSIONS: High plasma S100A12 levels on admission are associated with a poor functional outcome in patients with acute ischemic stroke.

Clinical significance of plasma VEGF value in ischemic stroke - research for biomarkers in ischemic stroke (REBIOS) study.

BACKGROUND: Vascular endothelial growth factor (VEGF) is a well-known molecule mediating neuronal survival and angiogenesis. However, its clinical significance in ischemic stroke is still controversial. The goal of this study was to examine the temporal profile of plasma VEGF value and its clinical significance in ischemic stroke with taking its subtypes into consideration. METHODS: We prospectively enrolled 171 patients with ischemic stroke and age- and gender-matched healthy subjects. The stroke patients were divided into 4 subtypes: atherothrombotic infarction (ATBI, n = 34), lacunar infarction (LAC, n = 45), cardioembolic infarction (CE, n = 49) and other types (OT, n = 43). Plasma VEGF values were measured as a part of multiplex immunoassay (Human MAP v1.6) and we obtained clinical information at 5 time points (days 0, 3, 7, 14 and 90) after the stroke onset. RESULTS: Plasma VEGF values were significantly higher in all stroke subtypes but OT than those in the controls throughout 90 days after stroke onset. There was no significant difference in the average VEGF values among ATBI, LAC, and CE. VEGF values were positively associated with neurological severity in CE patients, while a negative association was found in ATBI patients. After adjustment for possible confounding factors, plasma VEGF value was an independent predictor of poor functional outcome in CE patients. CONCLUSIONS: Although plasma VEGF value increases immediately after the stroke onset equally in all stroke subtypes, its significance in functional outcome may be different among the stroke subtypes.

Lead generation and optimization based on protein-ligand complementarity.

This work proposes a computational procedure for structure-based lead generation and optimization, which relies on the complementarity of the protein-ligand interactions. This procedure takes as input the known structure of a protein-ligand complex. Retaining the positions of the ligand heavy atoms in the protein binding site it designs structurally similar compounds considering all possible combinations of atomic species (N, C, O, CH(3), NH, etc). Compounds are ranked based on a score which incorporates energetic contributions evaluated using molecular mechanics force fields. This procedure was used to design new inhibitor molecules for three serine/threonine protein kinases (p38 MAP kinase, p42 MAP kinase (ERK2), and c-Jun N-terminal kinase 3 (JNK3)). For each enzyme, the calculations produce a set of potential inhibitors whose scores are in agreement with IC50 data and Ki values. Furthermore, the native ligands for each protein target, scored within the five top-ranking compounds predicted by our method, one of the top-ranking compounds predicted to inhibit JNK3 was synthesized and his inhibitory activity confirmed against ATP hydrolysis. Our computational procedure is therefore deemed to be a useful tool for generating chemically diverse molecules active against known target proteins.

In vitro selection of scFv and its production: an application of mRNA display and wheat embryo cell-free and E. coli cell production system.

Synthetic DNA libraries encoding human antibody V(L) and V(H) fragments were designed, constructed, and enriched using mRNA display. The enriched libraries were then combined to construct a scFv library for mRNA display. Sequencing revealed that 46% of the library coded for full-length scFvs. Considering the number of molecules used in mRNA display, the size of the library displayed was calculated to be >10(10). To verify this, we tried to isolate a scFv against human RANK. A scFv was successfully isolated in the sixth round of panning and was synthesized in wheat embryo cell-free (WE) and Escherichia coli cell systems. In the WE system, even though the production level was high, the product was almost soluble. However, in the E. coli system, it was over-produced as inclusion bodies. The inclusion bodies were successfully refolded and showed approximately the same binding affinity as the WE product. These results demonstrate that using mRNA display with synthetic libraries and WE and E. coli cell production systems, a system for in vitro selection and small- to large-scale production of scFvs has been established.

High-throughput sodium dodecyl sulfate polyacrylamide gel electrophoresis by linking conventional gel plates with automated liquid handler via Gel Adaptor.

The automation of SDS-PAGE required substantial investment. To lower this barrier, we devised a cost-effective, simple, and general method where samples are loaded on SDS-PAGE gels held by a newly devised "Gel Adaptor" on a general automated liquid handler. In this method, the liquid handler loads samples on the SDS-PAGE gels held at an angle of 80 degrees on the Gel Adaptor so that the samples can be vertically loaded on the narrow paths of the wells of the slanted gels. This method allows the automated liquid handler to load 144 samples on SDS-PAGE gels in approximately 10 min, greatly reducing the time and cost for high-throughput SDS-PAGE analyses.