0
Article ? AI-assigned paper type based on the abstract. Classification may not be perfect — flag errors using the feedback button. Tier 2 ? Original research — experimental, observational, or case-control study. Direct primary evidence. Sign in to save

Brahmaputra_MP_Analysis: Reproducible statistical and machine-learning pipeline for microplastic bioaccumulation in Brahmaputra River ichthyofauna

Zenodo (CERN European Organization for Nuclear Research) 2026
Papia Debnath, Raktim Sarmah, Sanayaima Singha, Silpisikha Deka, Utpal Kumar Das, Sarada Kanta Bhagabati, Harunur Rashid, Md. Mahmudunnabi Mithu, Yogeeta Dahal, Gaurav Paul, Rajdeep Dutta

Summary

Scientists studied fish from the Brahmaputra River (a major waterway in South Asia) and found that microplastic contamination in fish changes with the seasons—rainfall and river flow patterns affect how much plastic ends up in the fish people eat. Using this data, the researchers estimated how much microplastic people might consume through their diet, which matters because these fish are a food source for millions of people in the region. This research provides tools and methods other scientists can reuse to track plastic pollution in fish and better understand the risks to people who rely on these rivers for food.

Study Type Environmental

This repository contains the complete, reproducible Python analysis pipeline for the manuscript "Seasonal hydrology governs microplastic bioaccumulation and human dietary exposure in the ichthyofauna of the transboundary Brahmaputra River continuum." It includes cleaned input data (microplastic abundance, particle characterisation subsets, and monthly rainfall context) and code implementing non-parametric tests (Kruskal–Wallis, Dunn's post-hoc, Mann–Whitney U), a negative-binomial generalised linear model with a linear mixed-effects robustness check and station×season interaction test, PERMANOVA, exploratory Random Forest and gradient-boosting (SHAP) variable-importance analyses, tissue-partitioning and human dietary-intake estimation, and three ecotoxicological indices (Pollution Load Index, Polymer Hazard Index, Shannon diversity). Running python run_all.py regenerates every table and figure in the manuscript with a fixed random seed (42) for exact reproducibility. This work was supported by the Asia-Pacific Network for Global Change Research (APN), project CRRP2021-09MY-Rashid. v2 adds SEM surface texture quantification. pipeline (Supplimentary method Note SM1), including SDI analysis and Fig 4.

Share this paper