Netta Barak

Netta Barak

Auto-generated profile, based on publicly available information

Netta Barak

School of Computer Science and Engineering, The Hebrew University of Jerusalem, Jerusalem 91904, Israel

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 April 2021 in Science Translational Medicine
Verified
Journal Article
Open access
PDF file
NBNetta Barak
Netta Barak
RBRoni Ben-Ami
Roni Ben-Ami
TSTal Sido
Tal Sido
APAmir Perri
Amir Perri
ASAviad Shtoyer
Aviad Shtoyer
MYMoran Yassour
Moran Yassour
Adaptive sample pooling for SARS-CoV-2 testing based on positivity rates spares 76% of test reactions without compromising specificity or sensitivity.

15 June 2025 in bioRxiv
Verified
Preprint
Open access
PDF file
NBNetta Barak
Netta Barak
HBHaimasree Bhattacharya
Haimasree Bhattacharya
Francesco AsnicarFrancesco Asnicar
Francesco Asnicar
JSJaeyun Sung
Jaeyun Sung
NSNicola Segata
Nicola Segata
MYMoran Yassour
Moran Yassour
Predicting host phenotypes from shotgun metagenomic data is essential for translating microbiome research into clinical practice)Despite the development of numerous computational tools for this task, researchers often default to traditional machine learning methods such as Random Forest)This hesitancy to adopt newer methods stems from their complexity as well as the lack of standardized evaluations, as most tools are assessed on different datasets and compared against a limited set of methods)Here, we introduce LAMPP, a standardized benchmark for evaluating methods for predicting host phenotypes from gut metagenomic data)LAMPP features a diverse range of prediction tasks and enables consistent, comparative assessments across prediction tools)Our systematic evaluation of existing tools shows that classic machine learning methods (e.g., Random Forest) perform competitively, offering both ease of use and state-of-the-art results)At the same time, it demonstrates that microbiome-based phenotype prediction remains a challenging problem)By providing a consistent platform for ongoing evaluation and access to raw sequencing data, LAMPP motivates the development of novel prediction pipelines from raw sequencing data to phenotype prediction, including novel sample representation and data augmentation strategies)LAMPP is publicly available for ongoing benchmarking at https://lampp.yassourlab.com/ .

12 May 2026
Verified
Conference Paper
Closed access
NBNetta Barak
Netta Barak
Predicting host phenotypes from gut microbiome data is central to developing microbiome-based diagnostics and monitoring tools. Although many computational methods have been proposed, researchers frequently rely on traditional models such as Random Forest or Logistic Regression. This hesitancy to adopt newer methods stems from their complexity as well as the lack of standardized evaluations, as most tools are assessed on different datasets and compared against a limited set of methods. To address these challenges, we developed LAMPP — Live Assessment of Metagenomics-based tools for host Phenotype Prediction. LAMPP consists of multiple tasks covering five diverse phenotypes: inflammatory bowel disease, colorectal cancer, schizophrenia, delivery mode, and general health status. These tasks were designed to reflect real-world variation in dataset characteristics, including cohort size, class balance, technical heterogeneity, and train-test similarity (from within-cohort to cross-cohort scenarios). We evaluated a range of traditional and deep learning models, including Random Forest, XGBoost, Logistic Regression, SIAMCAT, DeepMicro, DEBIAS-M, TabPFN and fully connected neural networks. These models differ in their approaches to feature selection, normalization, and representation. Performance was measured using ROC AUC, and we additionally assessed runtime, resource usage, and usability. Our results show that prediction difficulty varies across tasks. Importantly, deep learning methods did not outperform traditional models. SIAMCAT, Random Forest, and XGBoost performed comparably, highlighting that accessible, established tools remain effective across diverse microbiome prediction scenarios. LAMPP and its leaderboard are available at https://lampp.yassourlab.com, providing a transparent platform for benchmarking emerging methods. By enabling consistent and reproducible comparisons, LAMPP fosters progress in microbiome-based phenotype prediction and supports the broader adoption of robust, accessible computational tools in microbiome research.