5,371 research outputs found
AI is a viable alternative to high throughput screening: a 318‑target study
Get PDFInformación suplementaria en: https://doi.org/10.1038/s41598-024-54655-z.The Atomwise AIMS Program está formado por más de 300 investigadores de distintos países.High throughput screening (HTS) is routinely used to identify bioactive small molecules. This requires physical compounds, which limits coverage of accessible chemical space. Computational approaches combined with vast on‑demand chemical libraries can access far greater chemical space, provided that the predictive accuracy is sufficient to identify useful molecules. Through the largest and most diverse virtual HTS campaign reported to date, comprising 318 individual projects, we demonstrate that our AtomNet® convolutional neural network successfully finds novel hits across every major therapeutic area and protein class. We address historical limitations of computational screening by demonstrating success for target proteins without known binders, high‑quality X‑ray crystal structures, or manual cherry‑picking of compounds. We show that the molecules selected by the AtomNet® model are novel drug‑like scaffolds rather than minor modifications to known bioactive compounds. Our empirical results suggest that computational methods can substantially replace HTS as the first step of small‑ molecule drug discovery
AI is a viable alternative to high throughput screening: a 318-target study
Get PDFData availability:
All data generated or analyzed during this study are included in this published article and its supplementary information files.Supplementary Information is available online at: https://www.nature.com/articles/s41598-024-54655-z#Sec15 .Correction to: Scientific Reports https://doi.org/10.1038/s41598-024-54655-z, published online 02 April 2024
The original version of this Article contained errors. The corrections are available online: The Atomwise AIMS Program. Author Correction: AI is a viable alternative to high throughput screening: a 318-target study. Sci Rep 14, 21579 (2024). DOI URL: https://doi.org/10.1038/s41598-024-70321-w (they are also available on the Corrigendum PDF file, below). .High throughput screening (HTS) is routinely used to identify bioactive small molecules. This requires physical compounds, which limits coverage of accessible chemical space. Computational approaches combined with vast on-demand chemical libraries can access far greater chemical space, provided that the predictive accuracy is sufficient to identify useful molecules. Through the largest and most diverse virtual HTS campaign reported to date, comprising 318 individual projects, we demonstrate that our AtomNet® convolutional neural network successfully finds novel hits across every major therapeutic area and protein class. We address historical limitations of computational screening by demonstrating success for target proteins without known binders, high-quality X-ray crystal structures, or manual cherry-picking of compounds. We show that the molecules selected by the AtomNet® model are novel drug-like scaffolds rather than minor modifications to known bioactive compounds. Our empirical results suggest that computational methods can substantially replace HTS as the first step of small-molecule drug discovery
AI is a viable alternative to high throughput screening : a 318-target study
Get PDFPublisher Copyright: © 2024. The Author(s).Peer reviewe
Author Correction: AI is a viable alternative to high throughput screening: a 318-target study
No full textAI is a viable alternative to high throughput screening: a 318-target study
No full textHigh throughput screening (HTS) is routinely used to identify bioactive small molecules. This requires physical compounds, which limits coverage of accessible chemical space. Computational approaches combined with vast on-demand chemical libraries can access far greater chemical space, provided that the predictive accuracy is sufficient to identify useful molecules. Through the largest and most diverse virtual HTS campaign reported to date, comprising 318 individual projects, we demonstrate that our AtomNet® convolutional neural network successfully finds novel hits across every major therapeutic area and protein class. We address historical limitations of computational screening by demonstrating success for target proteins without known binders, high-quality X-ray crystal structures, or manual cherry-picking of compounds. We show that the molecules selected by the AtomNet® model are novel drug-like scaffolds rather than minor modifications to known bioactive compounds. Our empirical results suggest that computational methods can substantially replace HTS as the first step of small-molecule drug discovery
AI is a viable alternative to high throughput screening : a 318-target study
No full textAbstract: High throughput screening ( HTS) is routinely used to identify bioactive small molecules. This requires physical compounds, which limits coverage of accessible chemical space. Computational approaches combined with vast on-demand chemical libraries can access far greater chemical space, provided that the predictive accuracy is sufficient to identify useful molecules. Through the largest and most diverse virtual HTS campaign reported to date, comprising 318 individual projects, we demonstrate that our AtomNet((R)) convolutional neural network successfully finds novel hits across every major therapeutic area and protein class. We address historical limitations of computational screening by demonstrating success for target proteins without known binders, high-quality X-ray crystal structures, or manual cherry-picking of compounds. We show that the molecules selected by the AtomNet((R)) model are novel drug-like scaffolds rather than minor modifications to known bioactive compounds. Our empirical results suggest that computational methods can substantially replace HTS as the first step of small-molecule drug discovery
Author Correction: AI is a viable alternative to high throughput screening: a 318-target study
No full textA case study in the evaluation of English training courses using a version of the CIPP model as an evaluative tool
Get PDFThis thesis presents an evaluative case study of the 20 English training courses offered in the Applied English Department (AED) of an Institute, given the pseudonym W.G, in southern Taiwan. No evaluation had been done since the AED had been set up and using Stufflebeam’s CIPP (Context, Input, Process and Product) evaluation model this research was carried out. The purpose of the research was to attempt, through the gathering of qualitative data from a variety of sources and using a variety of research instruments, an evaluation of the 20 English training courses which were designed for and taken by students who hoped, mainly, to become children's English language teachers. The courses were examined through four key components, namely, "course aims and objectives", "course contents and materials", "course conduct and teaching-learning process" and "assessment and student performance". Data were gathered through questionnaires, interviews and the review of existing documents and was obtained from current students, directors of the AED, instructors, alumni and employers of alumni. The resultant data served to present a comprehensive overview of the AED and the 20 English training courses and furnished evidence sufficient to allow for a number of recommendations for improvement and change to emerge. Fundamentally it is not clear that there is sufficient congruence of students needs and the courses offered. It emerged that the AED would probably benefit from a refocusing of student needs, a review of AED structures and governance, uniform syllabus design and presentation, a review of student feedback on instructor performance and a number of fundamental adjustments to the courses, in particular, their content, teaching methodology and assessment. Overall the AED had many positive aspects all of which could be built on and added to as the results of the data suggested. It emerged that the CIPP evaluation model has, in the educational context, a lot to commend it and this has been illustrated in this research. If followed carefully it covers all aspects and features of a program and provides a methodical, all-embracing design which can produce useful material for exploration and adoption if appropriate. It is in most cases a positive program enhancing exercise designed to develop rather than close existing programs
AI is a viable alternative to high throughput screening: a 318-target study
Get PDFHigh throughput screening (HTS) is routinely used to identify bioactive small molecules. This requires physical compounds, which limits coverage of accessible chemical space. Computational approaches combined with vast on-demand chemical libraries can access far greater chemical space, provided that the predictive accuracy is sufficient to identify useful molecules. Through the largest and most diverse virtual HTS campaign reported to date, comprising 318 individual projects, we demonstrate that our AtomNet® convolutional neural network successfully finds novel hits across every major therapeutic area and protein class. We address historical limitations of computational screening by demonstrating success for target proteins without known binders, high-quality X-ray crystal structures, or manual cherry-picking of compounds. We show that the molecules selected by the AtomNet® model are novel drug-like scaffolds rather than minor modifications to known bioactive compounds. Our empirical results suggest that computational methods can substantially replace HTS as the first step of small-molecule drug discovery
- …
