Transitioning GLYCAM-Web to a self-sustaining carbohydrate modeling service

点击翻译按钮获取中文摘要

Project Summary For over a decade, GLYCAM-Web (www.glycam.org) has provided the scientific community with a suite of online tools for simplifying the modeling of 3D structures of carbohydrates, glycoproteins, and their interactions with proteins. The highly plastic nature of oligosaccharides and the fact that they are typically branched means that they place unique demands on both software developers and users. This is exacerbated by their complex and archaic nomenclature. GLYCAM-web was developed by a team of people with a unique depth of expertise in carbohydrate chemistry and informatics, with the overarching goal of making carbohydrate modeling more accessible to scientists who either lack computational expertise, or whose familiarity with carbohydrates is limited. According to AWstats and in-house code GLYCAM-Web was visited by ~110,000 unique addresses and had ~11,000 PDB file downloads from 2016 to 2019. The downloads are usually associated with creating and viewing 3D structures of glycans and for offline modeling. We have included a detailed usage analysis in the Research Strategy to emphasize the significance of this resource to the NIGMS research community. The development of GLYCAM-Web has been primarily supported through an NIH NIGMS P41 grant held at the Complex Carbohydrate Research Center at the University of Georgia that has been continuously renewed since the 1990s. With the ending of the P41 program, we are not eligible for further support within the P41 program, and are seeking support that with enable us to transition GLYCAM-Web to a near self-sustaining resource whose operation costs are offset by new fee-for-service functionality targeting users who wish to perform molecular dynamics simulations, but who lack the expertise or infrastructure to do so. The decade- long development of GLYCAM-Web has seen the infrastructure grow and change with the advent of new architectures, code evolution, and increased security requirements, leading to a complex code base involving the C++, JAVA, and PYTHON languages. Thus a significant aspect of this proposal involves code refactoring to unify the code base, which will directly reduce maintenance costs. To achieve near self-sustainability, we have developed the four aims what will address the necessary improvements to code and infrastructure: usability, transferability, maintainability, and sustainability.