A post processing strategy to score and rank the annotation confidence of saponins in natural products by integrating MS2 spectral similarity and fragment interpretation.

人工智能 计算机科学 相似性(几何) 注释 化学 自然语言处理 片段(逻辑) 排名(信息检索) 一致性(知识库) 模式识别(心理学)
作者
Tong Xie,Jinjun Shan,Jun Jiang,Xia Zhao,Yu He,Wenjun Tong
出处
期刊:Journal of Pharmaceutical and Biomedical Analysis [Elsevier BV]
卷期号:204: 114291-
标识
DOI:10.1016/j.jpba.2021.114291
摘要

Tandem mass spectrometry-spectra-based annotation in natural products challenges a lot because of ambiguous structural characterization. It still lacks an efficiency method to score and rank the annotation confidence. Herein, we develop a novel approach to rank the annotation confidences of saponins. Annotations were accomplished according to fragmentation patterns. The corresponding diagnostic fragments and their abundances were recorded. Average abundances were taken as a reference spectrum, and the cosine similarity score (CSS) was calculated to measure how well the spectral matched. According to CSS values, statistic description for confidence levels can be effectively provided. Next, the fragment interpretation score (FIS) was proposed to investigate the deviators' characteristic fragmentation. FIS offset the effect from the deviators' unique fragments. Suspicious annotations involving low CSS and high FIS, may derived from the MS2 spectral background interferences or co-elution. Annotations with low CSS and FIS rank as low confidences, as these annotations need more attention. Using this method, novel saccharide sequences, specific fragmentation preferences, undistinguished precursors, even new structures can also be well traced. By proposed new scoring system, confidence evaluations can be ranked, resulting in significantly enhanced annotation reliability.

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
壮观雁易发布了新的文献求助10
刚刚
大模型的应助被哲学家采纳,获得10
刚刚
刚刚
CodeCraft的应助被ff06采纳,获得10
1秒前
1秒前
Zoe完成签到,获得积分10
2秒前
2秒前
2秒前
彭于晏的应助被cyl采纳,获得10
3秒前
3秒前
keyaner完成签到,获得积分10
3秒前
无花果的应助被zo采纳,获得10
4秒前
dawn发布了新的文献求助10
4秒前
5秒前
5秒前
冉义东完成签到,获得积分20
6秒前
大杨完成签到,获得积分10
6秒前
DREAMING发布了新的文献求助10
7秒前
霸气安筠发布了新的文献求助10
8秒前
8秒前
8秒前
XZM发布了新的文献求助10
8秒前
大气的师发布了新的文献求助20
8秒前
yys完成签到,获得积分20
9秒前
LIUDAN发布了新的文献求助10
10秒前
10秒前
11秒前
12秒前
霖29完成签到,获得积分10
12秒前
一见卿心完成签到,获得积分10
12秒前
chic完成签到 ,获得积分10
12秒前
笑一七完成签到,获得积分10
12秒前
13秒前
橙汁没泡泡完成签到,获得积分10
15秒前
15秒前
LIU发布了新的文献求助30
15秒前
17秒前
17秒前
18秒前
19秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
八旗制度の研究 600
Improving the wear resistance of vehicle gear oil by oil soluble LaF3 nanoparticles and typical additives of gear oil 500
Boundaries and Bodhisattvas 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8183694
求助须知:如何正确求助?哪些是违规求助? 9572963
关于积分的说明 21828306
捐赠科研通 7780864
什么是DOI,文献DOI怎么找? 3434057
关于科研通互助平台的介绍 2558257
邀请新用户注册赠送积分活动 2480541