A Method for Finding Metabolic Pathways Using Atomic Group Tracking

Abstract

A fundamental computational problem in metabolic engineering is to find pathways between compounds. Pathfinding methods using atom tracking have been widely used to find biochemically relevant pathways. However, these methods require the user to define the atoms to be tracked. This may lead to failing to predict the pathways that do not conserve the user-defined atoms. In this work, we propose a pathfinding method called AGPathFinder to find biochemically relevant metabolic pathways between two given compounds. In AGPathFinder, we find alternative pathways by tracking the movement of atomic groups through metabolic networks and use combined information of reaction thermodynamics and compound similarity to guide the search towards more feasible pathways and better performance. The experimental results show that atomic group tracking enables our method to find pathways without the need of defining the atoms to be tracked, avoid hub metabolites, and obtain biochemically meaningful pathways. Our results also demonstrate that atomic group tracking, when incorporated with combined information of reaction thermodynamics and compound similarity, improves the quality of the found pathways. In most cases, the average compound inclusion accuracy and reaction inclusion accuracy for the top resulting pathways of our method are around 0.90 and 0.70, respectively, which are better than those of the existing methods. Additionally, AGPathFinder provides the information of thermodynamic feasibility and compound similarity for the resulting pathways.

Nicotinamide attenuates the decrease in dendritic spine density in hippocampal primary neurons from 5xFAD mice, an Alzheimer’s disease animal model

“…For instance, the disturbances in metabolites of the glycolytic pathway (glucose-6-phosphate and glycerol-3-phosphate) and tricarboxylic acid (TCA) cycle (α-ketoglutarate, fumarate, and succinate) were identified in astrocytes derived from newborn 5xFAD mice [11], and pantethine treatment reduced the extent of metabolic perturbation and decreased the inflammatory processes in these astrocytes, indicating the role of altered brain energetics in the AD pathogenesis; metabolic profile analyses revealed region-specific metabolic changes in the hippocampus, cortex, cerebellum, and olfactory bulbs in APP/PS1 mice [12,13], and metabolomics signatures, including mitochondrial dysfunction and altered energy metabolism indicated by changes in nucleotide, TCA cycle, energy transfer, neurotransmitter, and amino acid metabolic pathways, were identified in APP/PS1 mice [14]; in addition, significant changes in metabolite compositions, including accumulation of fatty acids, alterations in phospholipids and acylcarnitines related to neural membrane degradation, and impaired energy management, were observed in the hippocampus and cortex in APP/PS1 mice [13]. Because the metabolic pathways are conserved through evolution [15,16], the metabolic signatures identified in AD mouse models could be directly translated into human studies [17]. Therefore, metabolomics screening in transgenic models could be useful for the understanding of the pathological mechanisms of AD.…”

Improving the organization and interactivity of metabolic pathfinding with precomputed pathways

“…When using the term 'conservation' we are specifically referring to retaining a subset of carbon atoms present in the starting compound through each biochemical transformation in the pathway through to the final target compound. Conserving some minimum number of carbon atoms throughout the pathway is done to ensure that the start compound is being utilized in production of the target compound, and this heuristic has been used by several existing metabolic pathfinding methods [9,10,12,17,19,44,45]. However, if the number of atoms conserved throughout the pathway is relatively small compared to the number of atoms in the start compound or target compound, the carbon conservation heuristic may not be as effective at preventing the search from finding infeasible pathways, since the search is still able to find pathways that break down the start compound to a much smaller compound (i.e., CO 2 ) then build that compound back up to produce the target compound.…”

“…Pathway Hunter Tool [14], MetabolicTinker [15], and GEM-Path [16] incorporate chemical similarity measures as heuristics to guide the metabolic search. Other methods track individual atoms, or groups of connected atoms in the case of AGPathFinder [17], from the start to target compound and aim to conserve a minimum number of atoms throughout the pathway [9,12,18] or maximize the number of atoms conserved [10,19]. Information on enzymatic reactions, like thermodynamic favorability ( G) [15,17,20] and enzyme efficiency and promiscuity [20,21], is also used to guide the search and rank pathways.…”

A review of parameters and heuristics for guiding metabolic pathfinding

“…This score accounts for five different atom types (carbon, oxygen, nitrogen, phosphorus, and sulfur), and each type of atom can be assigned a different weight. More recently, atom group tracking has been introduced by AGPathFinder [ 44 ]. Instead of tracking single atoms, this algorithm tracks groups of adjacent atoms connected by bonds.…”