Predicting the biodegradation of organic micropollutants

This study will teach us more about the potential biodegradability of problem substances. If we know whether, and how, the biodegradation of substances is possible, research can pave the way to controlling treatment processes by establishing the optimal conditions for removal.

Forecasting biological removal

A widely used method for assessing biodegradability is the OECD 309 method, which determines whether a substance is persistent when the half-life (DT50) exceeds 40 days in aerobic degradation conditions (KWR 2020.118). Software for predicting biodegradation (such as BIOWIN and BioTransformer) can indicate whether substances are biodegradable and what the possible metabolites are.

These tests and software are all based on specific conditions. However, many substances that do not comply with these conditions are therefore found to be persistent, even though they may be biodegradable. The aim of this study is to map out the biodegradability of these substances with the aim of possibly controlling treatment processes in the future so that these substances are removed as well.

Using text mining, QSARS and bioinformatics tools

KWR has conducted research in the past on predicting the removal of organic micropollutants with activated carbon and membrane processes, and during passage through the soil. In this study, we aim to predict whether known and unknown persistent substances are actually biodegradable. In addition, we wish to establish an overview of the conditions, and the microorganisms and enzymes involved in biodegradation. With the help of text mining in the literature, and target and non-target screening data from the drinking water companies, we will review which persistent problem substances the drinking and wastewater companies are faced with and how they can be biodegraded. We will then be able to investigate why these substances are persistent by correlating this knowledge with the properties of the substances, in other words by means of clustering into substance groups on the basis of the chemical fingerprints of the substances, and applying Quantitative Structure-Activity Relationships (QSARs). We will also look at the microorganisms and enzymes involved in this degradation using bioinformatics tools. This will allow us to discover biodegradation pathways, bacteria and/or enzymes and compare them using QSAR clustering.

Moving towards prediction models

With the steps described here, we may be able to produce new prediction models that include the conditions and enzymes needed to degrade the substance. Many prediction tools and models have been developed in recent years that can identify problem substances. We want to determine which tool is the most effective and useful. The application and development of these computational tools to describe biodegradability while combining the different analyses constitutes a new approach. We hope that this exploration of the combined use of bioinformatics and cheminformatics in hydroinformatics will produce new insights and research lines.