Machine Learning of Gut Microbiome Composition for Diagnostic Classification of Inflammatory Bowel Diseases

作者
Ishan Manandhar,Ahmad Alimadadi,Sachin Aryal,Patricia B. Munroe,Bina Joe,Xi Ran Cheng
出处
期刊:The FASEB Journal [Wiley]
卷期号:35 (S1) 被引量:2
标识
DOI:10.1096/fasebj.2021.35.s1.04113
摘要

Crohn's disease (CD) and ulcerative colitis (UC) are two major subtypes of inflammatory bowel disease (IBD) characterized by chronic inflammation of the gastrointestinal tract. Despite various clinical approaches being available for diagnosing IBD, such as blood tests and colonoscopy, misdiagnosis of IBD and its subtypes occurs frequently, thus there is a clinical need to improve diagnosis of this condition. Since dysregulated gut microbiota is reported in IBD patients, we hypothesized that gut microbiome data could be used to develop an artificial intelligence‐based strategy for detection of IBD and diagnostic differentiation of CD and UC (Figure 1). First, five supervised machine learning (ML) approaches, including random forest (RF), decision tree, elastic net, support vector machine and neural networks, were used to analyze fecal 16S metagenomics data of 729 IBD patients and 700 non‐IBD controls collected from the American Gut Project. Fifty significantly differential bacterial taxa were identified between the IBD and non‐IBD groups. ML classifications, trained with these 50 taxonomic features, achieved a testing AUC (area under the receiver operating characteristics curve) of ~0.80 (Figure 2A) using the trained RF model. Next, instead of bacterial taxa, we tested if operational taxonomic units (OTUs) could be used as ML features for diagnosing IBD. RF modeling of the top high‐variance OTU features achieved an improved AUC of ~0.82 (Figure 2B). Lastly, we tested the capability of the RF model to distinguish between CD and UC using fecal 16S metagenomics data of 331 CD and 141 UC subjects. A total of 117 significantly differential bacterial taxa were identified and the RF model trained with these bacterial features achieved a testing AUC of ~0.91 (Figure 2C). Furthermore, the RF model trained with the top high‐variance OTU features achieved an improved AUC of ~0.92 (Figure 2C). In summary, our study demonstrates the promising potential of artificial intelligence via supervised ML modeling of large‐scale gut microbiome data for detection and diagnostic classification of IBD.

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
小陈完成签到,获得积分10
刚刚
刚刚
火鸡味锅巴完成签到 ,获得积分10
刚刚
淡定采波完成签到,获得积分10
刚刚
落后的嚣完成签到,获得积分10
刚刚
GQ完成签到 ,获得积分10
1秒前
Preveil完成签到,获得积分10
1秒前
小周发布了新的文献求助10
1秒前
娜娜子完成签到 ,获得积分10
2秒前
xxszyb完成签到,获得积分10
3秒前
elysia完成签到,获得积分10
3秒前
3秒前
小蘑菇的应助被XY采纳,获得10
3秒前
ZZZ完成签到 ,获得积分10
4秒前
矜持完成签到,获得积分10
4秒前
雪糕发布了新的文献求助10
4秒前
大东东完成签到,获得积分10
4秒前
庾傀斗发布了新的文献求助10
4秒前
陈丰滢完成签到,获得积分10
4秒前
4秒前
颜夜关注了科研通微信公众号
5秒前
望望旺仔牛奶完成签到,获得积分10
6秒前
oldjeff完成签到,获得积分10
6秒前
Verna完成签到,获得积分10
7秒前
熊火焰完成签到,获得积分10
7秒前
zjxnq完成签到 ,获得积分10
7秒前
科研大大完成签到 ,获得积分10
7秒前
jzj完成签到 ,获得积分0
7秒前
风息完成签到,获得积分10
8秒前
鱼大大完成签到,获得积分10
8秒前
浮生若梦完成签到,获得积分10
8秒前
小蒋完成签到,获得积分10
9秒前
turnsole完成签到,获得积分10
9秒前
吉不二完成签到,获得积分10
9秒前
bobo完成签到,获得积分10
9秒前
藜誌完成签到,获得积分10
10秒前
狐狸的贝完成签到,获得积分10
11秒前
abab小王完成签到,获得积分10
11秒前
飞飞完成签到 ,获得积分10
11秒前
zhang完成签到,获得积分10
11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 1500
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
John McDowells Bildungstheorie 800
八旗制度の研究 600
Analyses de reconnaissances de dettes passées devant les échevins d'Ypres (1249-1291) 600
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 8159288
求助须知:如何正确求助?哪些是违规求助? 9554702
关于积分的说明 21720900
捐赠科研通 7748911
什么是DOI,文献DOI怎么找? 3427184
关于科研通互助平台的介绍 2552111
邀请新用户注册赠送积分活动 2471479