A primer on machine learning techniques for genomic applications

Monaco, A.; Pantaleo, Ester; Amoroso, Nicola; Lacalamita, Antonio; Giudice, Claudio Lo; Fonzino, Adriano; Fosso, Bruno; Picardi, Ernesto; Tangaro, Sabina; Pesole, Graziano; Bellotti, Roberto

doi:10.1016/j.csbj.2021.07.021

Cited by 21 publications

(8 citation statements)

References 77 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To make the dataset more balanced, a data augmentation algorithm, SMOTE (Synthetic Minority Over-sampling Technique) 44 , was adopted. As a classification algorithm, random forest was chosen using the default parameters of the Scikit-learn python package 40 due to its efficiency in handling datasets with a high number of features 45 . The final output was obtained as the average of the three probability values and the associated class was obtained from the probability value by imposing the classic threshold of 0.5.…”

Section: Resultsmentioning

confidence: 99%

Microbiome Preterm Birth DREAM Challenge: Crowdsourcing Machine Learning Approaches to Advance Preterm Birth Research

Golob

Oskotsky

Tang

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Globally, every year about 11% of infants are born preterm, defined as a birth prior to 37 weeks of gestation, with significant and lingering health consequences. Multiple studies have related the vaginal microbiome to preterm birth. We present a crowdsourcing approach to predict: (a) preterm or (b) early preterm birth from 9 publicly available vaginal microbiome studies representing 3,578 samples from 1,268 pregnant individuals, aggregated from raw sequences via an open-source tool, MaLiAmPi. We validated the crowdsourced models on novel datasets representing 331 samples from 148 pregnant individuals. From 318 DREAM challenge participants we received 148 and 121 submissions for our two separate prediction sub-challenges with top-ranking submissions achieving bootstrapped AUROC scores of 0.69 and 0.87, respectively. Alpha diversity, VALENCIA community state types, and composition (via phylotype relative abundance) were important features in the top performing models, most of which were tree based methods. This work serves as the foundation for subsequent efforts to translate predictive tests into clinical practice, and to better understand and prevent preterm birth.

show abstract

Section: Resultsmentioning

confidence: 99%

Microbiome Preterm Birth DREAM Challenge: Crowdsourcing Machine Learning Approaches to Advance Preterm Birth Research

Golob

Oskotsky

Tang

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…As a classification algorithm, random forest was chosen using the default parameters of the Scikit-learn python package 46 due to its efficiency in handling datasets with a high number of features. 75 The final output was obtained as the average of the three probability values and the associated class was obtained from the probability value by imposing the classic threshold of 0.5. The prediction model achieved an AUROC of 0.64 and an AUPRC of 0.48 on the Dream Challenge validation dataset.…”

Section: Methodsmentioning

confidence: 99%

Microbiome preterm birth DREAM challenge: Crowdsourcing machine learning approaches to advance preterm birth research

Golob,

Oskotsky,

Tang

et al. 2024

Cell Reports Medicine

Self Cite

View full text Add to dashboard Cite

“…Throughout the chapter, there have been examples showcasing the use of ML in combination with IoT devices in the biotechnology Industry 5.0. These models are used for an array of solutions in the biotechnology industry, including predictive analysis for crop yield [53], classification of genome data [54], enhanced manufacturing of biotechnology products [55], management of biotechnology labs [56], and more. Like other components of Industry 5.0, the ML models are also vulnerable.…”

Section: The Role Of Mlsecops In the Biotechnology Industry 50mentioning

confidence: 99%

Integrating MLSecOps in the Biotechnology Industry 5.0

Pervez,

J. Titus

2024

The Role of Cybersecurity in the Industry 5.0 Era [Working Title]

View full text Add to dashboard Cite

Biotechnology Industry 5.0 is advancing with the integration of cutting-edge technologies such as Machine Learning (ML), the Internet of Things (IoT), and cloud computing. It is no surprise that an industry that utilizes data from customers and can alter their lives is a target of a variety of attacks. This chapter provides a perspective on how Machine Learning Security Operations (MLSecOps) can help secure the biotechnology Industry 5.0. The chapter provides an analysis of the threats in the biotechnology Industry 5.0 and how ML algorithms can help secure with industry best practices. This chapter explores the scope of MLSecOps in the biotechnology Industry 5.0, highlighting how crucial it is to comply with current regulatory frameworks. With biotechnology Industry 5.0 developing innovative solutions in healthcare, supply chain management, biomanufacturing, pharmaceutical sectors, and more, the chapter also discusses the MLSecOps best practices that industry and enterprises should follow while also considering ethical responsibilities. Overall, the chapter provides a discussion of how to integrate MLSecOps into the design, deployment, and regulation of the processes in the biotechnology Industry 5.0.

show abstract

A primer on machine learning techniques for genomic applications

Abstract: Graphical abstract

Cited by 21 publications

References 77 publications

Microbiome Preterm Birth DREAM Challenge: Crowdsourcing Machine Learning Approaches to Advance Preterm Birth Research

Microbiome Preterm Birth DREAM Challenge: Crowdsourcing Machine Learning Approaches to Advance Preterm Birth Research

Microbiome preterm birth DREAM challenge: Crowdsourcing machine learning approaches to advance preterm birth research

Integrating MLSecOps in the Biotechnology Industry 5.0

Contact Info

Product

Resources

About