转化(遗传学)
化学
计算机科学
万维网
生物化学
基因
作者
Jibao Liu,Bei Zhang,Qing‐Long Fu,Toshihiro Isobe,Rongjun Gao,Yuansong Wei,Eunsang Kwon,Zhineng Hao,Wei An,Qi Rong,Manabu Fujii
标识
DOI:10.1021/acs.estlett.5c00284
摘要
Dissolved fractions of organic matter in natural and anthropogenic sources, also known as dissolved organic matter (DOM), play crucial roles in natural and engineering processes. Application of the Fourier transform ion cyclotron resonance mass spectrometer (FT-ICR MS) has expanded dramatically for the molecular characterization of DOM in various research communities. However, data processing/mining on the detected tens of thousands of molecular formulas is complex. Here, we introduced a browser-based webtool (MoleTrans) for flexible and comprehensive postanalysis on a formula-assigned data set from FT-ICR MS. MoleTrans includes the classical but fundamental analysis/visualization techniques (e.g., chemodiversity, multivariate statistics, van Krevelen diagrams, formula composition plots, Kendrick Mass Defect (KMD) based homologous series analysis). Users can explore the putative molecular transformation relations in single or multiple samples using a Paired Mass Distance (PMD) network in a flexible manner (e.g., user-defined mass errors and reference transformation groups, etc.). This unique tool also supports the machine learning (ML) workflow (data processing and training models) for explaining the molecular transformation behaviors between samples (e.g., exploring the disappeared, resistant, and newly appeared formulas during the transformation). Therefore, MoleTrans provides the opportunity to unravel the molecular fingerprint in DOM mixtures in a comprehensive, flexible, and reproducible way.
科研通智能强力驱动
Strongly Powered by AbleSci AI