Knowledge-Inclusive Adaptive Physics-Informed Neural Network for Microbial Interaction Modelling

Asela Hevapathige; Rajith Vidanaarachchi; Ravisha Rupasinghe; Sachith Seneviratne; Saman Halgamuge; Sen-Lin Tang

arxiv: 2606.07686 · v1 · pith:6TWYVQBEnew · submitted 2026-06-05 · 💻 cs.LG · cs.AI

Knowledge-Inclusive Adaptive Physics-Informed Neural Network for Microbial Interaction Modelling

Ravisha Rupasinghe , Rajith Vidanaarachchi , Asela Hevapathige , Sachith Seneviratne , Sen-Lin Tang , Saman Halgamuge This is my paper

Pith reviewed 2026-06-27 22:32 UTC · model grok-4.3

classification 💻 cs.LG cs.AI

keywords physics-informed neural networksgeneralized Lotka-Volterramicrobial interaction modelingknowledge integrationmetagenomics literaturenetwork inferenceadaptive PINNcommunity dynamics

0 comments

The pith

A physics-informed neural network that adds literature text and network structure to equations discovers better parameters for microbial interactions.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a framework that extends physics-informed neural networks beyond differential equations to also incorporate auxiliary knowledge from peer-reviewed metagenomics literature as text and from explicit microbial interaction networks. This enriched model is applied to the generalized Lotka-Volterra equations to discover parameters from both experimental abundance measurements and the added knowledge sources. A sympathetic reader would care because current PINN approaches for microbial communities rely only on measurements and therefore miss biological context on external influences, while the new method reports clear accuracy gains on both real and simulated human- and plant-associated datasets. The work further shows that the inferred networks align with ecological roles already documented in the literature.

Core claim

The central claim is that enriching generalized Lotka-Volterra parameters with text-based knowledge from metagenomics literature (capturing external influences) and with network-based structural knowledge (explicitly modeling interactions), then combining these with experimental data through a data-driven integration approach inside an adaptive PINN, produces more accurate microbial community models than existing methods that use only measurements.

What carries the argument

The knowledge-inclusive adaptive Physics-Informed Neural Network that performs data-driven integration of text and network knowledge into generalized Lotka-Volterra parameter discovery.

If this is right

The framework improves over prior state-of-the-art methods by up to 53 percent even when no additional knowledge is supplied.
Knowledge addition produces further gains of up to 23 percent in Bray-Curtis dissimilarity accuracy and 47 percent in R-squared.
The model infers microbial interaction networks that can be checked against ecological roles documented in the literature.
The same approach is demonstrated on both human-associated and plant-associated communities using real and simulated abundance data.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same knowledge-integration pattern could be tested on other equation-based models in ecology or systems biology where supporting text sources exist.
If the accuracy gains hold, the method might reduce the amount of experimental data needed to reach a given prediction quality for microbial dynamics.
Extending the framework to include time-varying external factors described in the literature could address limitations of the base generalized Lotka-Volterra form.

Load-bearing premise

Auxiliary knowledge extracted from peer-reviewed metagenomics literature can be integrated into the PINN without introducing inconsistencies or biases that distort the discovered parameters.

What would settle it

On the paper's real or simulated datasets, adding the literature-derived knowledge produces no gain or a loss in Bray-Curtis dissimilarity accuracy or R-squared compared with the version that uses only experimental data and the base PINN.

read the original abstract

Physics-Informed Neural Network (PINN) is a way of including knowledge in the form of equations in Machine Learning methods. Beyond equations, knowledge exists in other forms, such as text and network structure. While existing PINN-based approaches discover equation parameters from data, they rely solely on experimental measurements. We propose a new PINN framework that enriches parameter discovery by incorporating auxiliary knowledge sources. We instantiate our framework for microbiology, where generalised Lotka-Volterra (gLV) serves as a biological foundation for modelling microbial communities. We demonstrate that incorporating knowledge improves microbial community modelling. Our framework enriches the gLV parameters using peer-reviewed metagenomics literature, as text provides biological context on external influences that gLV alone cannot capture. We combine this knowledge with experimental measurements of microbial abundance using a data-driven integration approach. We integrate network-based structural knowledge by explicitly modelling microbial interactions. Our knowledge-inclusive framework infers microbial networks, revealing ecological insights. We validate these findings against ecological roles documented in the literature. We evaluate on real and simulated datasets spanning human- and plant-associated microbial communities. Our framework improves over the state-of-the-art by up to 53%, even without knowledge. Knowledge addition yields gains of up to 23% in Bray-Curtis Dissimilarity-based accuracy and 47% in $\mathrm{R}^2$.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper adds text from metagenomics literature and explicit network modeling to an adaptive PINN on gLV equations, with claimed gains, but the integration protocol stays too vague to judge if the improvements are real or curated.

read the letter

This paper's main contribution is a method to fold text from the metagenomics literature and network structure into an adaptive PINN on top of the generalized Lotka-Volterra equations for modeling microbial interactions.

It improves on standard PINN baselines by up to 53% even without the extra knowledge, and the added sources give further gains of 23% in Bray-Curtis dissimilarity and 47% in R-squared on both simulated and real datasets from human and plant-associated communities. The fact that they check the inferred networks against known ecological roles from the literature is useful for showing the outputs make biological sense. The framework seems to treat the text as auxiliary context that gLV alone misses, like external influences, and combines it with abundance measurements.

The integration of the text knowledge is the weakest part. The description stays at the level of 'data-driven integration approach' and 'enriches the gLV parameters using peer-reviewed metagenomics literature,' without spelling out how individual statements get turned into specific loss terms or parameter bounds. If that step involves manual selection or unvalidated extraction, the reported improvements could be inflated by curation bias rather than the method. The stress-test note flags exactly this, and the abstract alone does not resolve it. No error bars or detailed validation procedures are mentioned either.

On the positive side, the base improvement without knowledge suggests the adaptive PINN itself is doing something useful, and the network modeling part for inferring interactions is a reasonable extension.

This is aimed at researchers in microbial ecology or computational biology who work with PINNs and want to incorporate more than just differential equations. A reader from pure ML might not get much without the domain context.

I would recommend sending it for peer review because the core idea has enough novelty and the claims are specific enough to be checked by referees who can ask for the exact integration protocol and reproducibility details. It is not ready as is, but it deserves the time.

Referee Report

1 major / 2 minor

Summary. The manuscript proposes a Knowledge-Inclusive Adaptive Physics-Informed Neural Network (KI-APINN) framework that extends PINNs for microbial community modeling based on the generalized Lotka-Volterra (gLV) equations. It incorporates auxiliary knowledge from peer-reviewed metagenomics literature (as text) and network structure in addition to experimental abundance data, using a data-driven integration approach to enrich parameter discovery. The authors report that the framework improves over state-of-the-art by up to 53% even without knowledge, with further gains of up to 23% in Bray-Curtis dissimilarity accuracy and 47% in R² from knowledge addition; inferred networks are validated against documented ecological roles, with evaluation on real and simulated datasets from human- and plant-associated communities.

Significance. If the knowledge integration is shown to be reproducible and free of curation bias, the work could meaningfully advance PINN applications in systems biology by demonstrating systematic inclusion of non-equation knowledge sources. Strengths include the combination of simulated and real data plus literature validation of inferred interactions. No mention of open code or machine-checked proofs is present.

major comments (1)

[Methods] Methods (knowledge integration subsection): the protocol for extracting statements from peer-reviewed metagenomics literature and mapping them to auxiliary constraints, priors, or loss terms in the adaptive PINN is described only generically as a 'data-driven integration approach.' This is load-bearing for the central claim, because the headline gains attributed to knowledge addition (23% Bray-Curtis, 47% R²) cannot be assessed for robustness without an explicit, reproducible extraction and validation procedure that rules out selective curation or introduced inconsistencies in gLV parameter discovery.

minor comments (2)

[Abstract] Abstract and Results: performance claims are given as 'up to' maxima without accompanying error bars, number of replicates, data-split details, or statistical tests, which reduces clarity on the reliability of the reported improvements.
Notation: the distinction between the base adaptive PINN and the knowledge-augmented version should be made explicit in equations and algorithm descriptions to avoid reader confusion.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback. We address the single major comment below and commit to a revision that strengthens the reproducibility of the knowledge-integration protocol.

read point-by-point responses

Referee: [Methods] Methods (knowledge integration subsection): the protocol for extracting statements from peer-reviewed metagenomics literature and mapping them to auxiliary constraints, priors, or loss terms in the adaptive PINN is described only generically as a 'data-driven integration approach.' This is load-bearing for the central claim, because the headline gains attributed to knowledge addition (23% Bray-Curtis, 47% R²) cannot be assessed for robustness without an explicit, reproducible extraction and validation procedure that rules out selective curation or introduced inconsistencies in gLV parameter discovery.

Authors: We agree that the current description is insufficiently explicit. In the revised manuscript we will add a dedicated subsection titled 'Knowledge Extraction and Integration Protocol' that details: (1) the literature search strategy and inclusion criteria, (2) the sentence-level extraction rules and entity mapping to microbial taxa and interaction signs, (3) the exact transformation of extracted statements into auxiliary loss terms or parameter priors within the adaptive PINN objective, and (4) the validation steps used to check consistency with the gLV structure. We will also release the extraction scripts and the curated statement set as supplementary material to allow independent reproduction. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation self-contained

full rationale

Abstract and framework description present empirical gains from adding external literature-derived knowledge and network structure to a PINN-gLV model, validated against independent ecological literature. No equations, fitted parameters renamed as predictions, or self-citation chains are exhibited that reduce outputs to inputs by construction. Central claims rest on external benchmarks and data integration rather than definitional equivalence or load-bearing self-references.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Insufficient information in the abstract to identify specific free parameters, axioms, or invented entities; no details on how knowledge is encoded or any new postulated quantities.

pith-pipeline@v0.9.1-grok · 5798 in / 1177 out tokens · 32445 ms · 2026-06-27T22:32:46.122788+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

40 extracted references · 5 canonical work pages · 1 internal anchor

[1]

Journal of Computational physics378, 686–707 (2019)

Raissi, M., Perdikaris, P., Karniadakis, G.E.: Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational physics378, 686–707 (2019)

2019
[2]

arXiv preprint arXiv:2509.02282 (2025)

Fontanarrosa, P., Barnes, C.P.: Exploring accuracy and uncertainty quantification in physics-informed neural networks for inferring microbial community dynamics. arXiv preprint arXiv:2509.02282 (2025)

work page arXiv 2025
[3]

bioRxiv (2026) https://doi.org/10.64898/2026.04.29.721522 https://www.biorxiv.org/content/early/2026/05/02/2026.04.29.721522.full.pdf

Gamage, C.J., Xia, Y., Rupasinghe, R., Senevirathne, S., Senanayake, D., Malepathirana, T., Hevapathige, A., Corbett, M., O’Brien, T.J., Petrou, S., Berkovic, S.F., Scheffer, I.E., Gecz, J., Bahlo, M., Bennett, M.F., Halgamuge, S.K.: Knowledge inclusive machine learning for disease gene prioritisation. bioRxiv (2026) https://doi.org/10.64898/2026.04.29.72...

work page doi:10.64898/2026.04.29.721522 2026
[4]

American Society for Microbiology 1752 N St., NW, Washington, DC (2016)

Blaser, M.J., Cardon, Z.G., Cho, M.K., Dangl, J.L., Donohue, T.J., Green, J.L., Knight, R., Maxon, M.E., Northen, T.R., Pollard, K.S., et al.: Toward a predic- tive understanding of Earth’s microbiomes to address 21st century challenges. American Society for Microbiology 1752 N St., NW, Washington, DC (2016)

2016
[5]

Nature metabolism1(1), 34–46 (2019)

Cani, P.D., Van Hul, M., Lefort, C., Depommier, C., Rastelli, M., Everard, A.: Microbial regulation of organismal energy homeostasis. Nature metabolism1(1), 34–46 (2019)

2019
[6]

Cell148(6), 1258–1270 (2012)

Clemente, J.C., Ursell, L.K., Parfrey, L.W., Knight, R.: The impact of the gut microbiota on human health: an integrative view. Cell148(6), 1258–1270 (2012)

2012
[7]

Advanced drug delivery reviews210, 115332 (2024) 29

Cho, Y.S., Han, K., Xu, J., Moon, J.J.: Novel strategies for modulating the gut microbiome for cancer therapy. Advanced drug delivery reviews210, 115332 (2024) 29

2024
[8]

Bergey, D., Hendricks: Bergey’s Manual of Determinative Bacteriology, 9th edn. (1994)

1994
[9]

Nature Reviews Microbiology10(8), 538–550 (2012)

Faust, K., Raes, J.: Microbial interactions: from networks to models. Nature Reviews Microbiology10(8), 538–550 (2012)

2012
[10]

Elife11, 73870 (2022)

Baranwal, M., Clark, R.L., Thompson, J., Sun, Z., Hero, A.O., Venturelli, O.S.: Recurrent neural networks enable design of multifunctional synthetic human gut microbiome dynamics. Elife11, 73870 (2022)

2022
[11]

BMC Molecular and Cell Biology21(Suppl 1), 34 (2020)

Vidanaarachchi, R., Shaw, M., Tang, S.-L., Halgamuge, S.: Imparo: inferring microbial interactions through parameter optimisation. BMC Molecular and Cell Biology21(Suppl 1), 34 (2020)

2020
[12]

In: Forty-first International Conference on Machine Learning (2024)

Ruaud, A., Sancaktar, C., Bagatella, M., Ratzke, C., Martius, G.: Modelling microbial communities with graph neural networks. In: Forty-first International Conference on Machine Learning (2024)

2024
[13]

Msystems8(3), 00757–22 (2023)

Dedrick, S., Warrier, V., Lemon, K.P., Momeni, B.: When does a lotka-volterra model represent microbial interactions? insights from in vitro nasal bacterial communities. Msystems8(3), 00757–22 (2023)

2023
[14]

PLoS computational biology16(5), 1007917 (2020)

Joseph, T.A., Shenhav, L., Xavier, J.B., Halperin, E., Pe’er, I.: Compositional lotka-volterra describes microbial dynamics in the simplex. PLoS computational biology16(5), 1007917 (2020)

2020
[15]

bioRxiv (2024)

Ma, C., Liu, S., Koslicki, D.: Metagenomickg: a knowledge graph for metagenomic applications. bioRxiv (2024)

2024
[16]

Genome biology12(5), 50 (2011)

Caporaso, J.G., Lauber, C.L., Costello, E.K., Berg-Lyons, D., Gonzalez, A., Stombaugh, J., Knights, D., Gajer, P., Ravel, J., Fierer, N.,et al.: Moving pictures of the human microbiome. Genome biology12(5), 50 (2011)

2011
[17]

Microbiology spectrum10(5), 01803–22 (2022)

Wang, C.-W., Michelle Wong, J.-W., Yeh, S.-S., Eric Hsieh, Y., Tseng, C.-H., Yang, S.-H., Tang, S.-L.: Soil bacterial community may offer solutions for ginger cultivation. Microbiology spectrum10(5), 01803–22 (2022)

2022
[18]

Nature communications12(1), 3254 (2021)

Clark, R.L., Connors, B.M., Stevenson, D.M., Hromada, S.E., Hamilton, J.J., Amador-Noguez, D., Venturelli, O.S.: Design of synthetic human gut microbiome assembly and butyrate production. Nature communications12(1), 3254 (2021)

2021
[19]

Microbiome6(1), 193 (2018)

Grier, A., McDavid, A., Wang, B., Qiu, X., Java, J., Bandyopadhyay, S., Yang, H., Holden-Wiltse, J., Kessler, H.A., Gill, A.L.,et al.: Neonatal gut and respiratory microbiota: coordinated development through time and space. Microbiome6(1), 193 (2018)

2018
[20]

Microbiology Spectrum12(10), 04109–23 (2024)

Karwowska, Z., Szczerbiak, P., Kosciolek, T.: Microbiome time series data reveal 30 predictable patterns of change. Microbiology Spectrum12(10), 04109–23 (2024)

2024
[21]

Nature Ecology & Evolution5(6), 747–756 (2021)

Averill, C., Werbin, Z.R., Atherton, K.F., Bhatnagar, J.M., Dietze, M.C.: Soil microbiome predictability increases with spatial and taxonomic scale. Nature Ecology & Evolution5(6), 747–756 (2021)

2021
[22]

MBio13(1), 03161–21 (2022)

Armour, C.R., Top¸ cuo˘ glu, B.D., Garretto, A., Schloss, P.D.: A goldilocks princi- ple for the gut microbiome: taxonomic resolution matters for microbiome-based classification of colorectal cancer. MBio13(1), 03161–21 (2022)

2022
[23]

Frontiers in Microbiology15, 1442208 (2024)

Cho, G., Kim, D.-R., Kwak, Y.-S.: Ecological shifts in soil microbiota and root rot disease progress during ginseng monoculture. Frontiers in Microbiology15, 1442208 (2024)

2024
[24]

ISME communications1(1), 31 (2021)

Reitmeier, S., Hitch, T.C., Treichel, N., Fikas, N., Hausmann, B., Ramer-Tait, A.E., Neuhaus, K., Berry, D., Haller, D., Lagkouvardos, I.,et al.: Handling of spurious sequences affects the outcome of high-throughput 16s rrna gene amplicon profiling. ISME communications1(1), 31 (2021)

2021
[25]

https://pubmed.ncbi.nlm.nih.gov/

National Library of Medicine: PubMed. https://pubmed.ncbi.nlm.nih.gov/. Accessed: 2025-09-29

2025
[26]

Minilm: Deep self-attention distillation for task-agnostic compression of pre-trained transformers, 2020

Wang, X., Zhang, F.W., Wang, M., Li, J., Zhou, M.: Minilm: Deep self- attention distillation for task-agnostic compression of pretrained transformers. arXiv preprint arXiv:2002.10957 (2020)

work page arXiv 2002
[27]

In: Proceedings of NAACL-HLT (2019)

Devlin, J., Chang, M.-W., Lee, K., Toutanova, K.: Bert: Pre-training of deep bidirectional transformers for language understanding. In: Proceedings of NAACL-HLT (2019)

2019
[28]

bioRxiv (2024) https://doi.org/10.1101/2024.06.30.601456 https://www.biorxiv.org/content/early/2024/07/02/2024.06.30.601456.full.pdf

Vidanaarachchi, R., Tang, S.L., Halgamuge, S.: Copr: Collective pattern recognition—a framework for microbial community activ- ity analysis. bioRxiv (2024) https://doi.org/10.1101/2024.06.30.601456 https://www.biorxiv.org/content/early/2024/07/02/2024.06.30.601456.full.pdf

work page doi:10.1101/2024.06.30.601456 2024
[29]

Anaerobe85, 102819 (2024)

Shin, J.H., Tillotson, G., MacKenzie, T.N., Warren, C.A., Wexler, H.M., Gold- stein, E.J.: Bacteroides and related species: The keystone taxa of the human gut microbiota. Anaerobe85, 102819 (2024)

2024
[30]

Cell host & microbe31(4), 485–499 (2023)

Culp, E.J., Goodman, A.L.: Cross-feeding in the gut microbiome: ecology and mechanisms. Cell host & microbe31(4), 485–499 (2023)

2023
[31]

MBio5(3), 10–1128 (2014)

Coyne, M.J., Zitomersky, N.L., McGuire, A.M., Earl, A.M., Comstock, L.E.: Evi- dence of extensive dna transfer between bacteroidales species within the human gut. MBio5(3), 10–1128 (2014)

2014
[32]

Nature Biotechnology 41(10), 1424–1433 (2023)

Huang, Y., Sheth, R.U., Zhao, S., Cohen, L.A., Dabaghi, K., Moody, T., Sun, Y., 31 Ricaurte, D., Richardson, M., Velez-Cortes, F.,et al.: High-throughput micro- bial culturomics using automation and machine learning. Nature Biotechnology 41(10), 1424–1433 (2023)

2023
[33]

Agronomy13(5), 1392 (2023)

Acu˜ na, J.J., Rilling, J.I., Inostroza, N.G., Manquian, J., Zhang, Q., Gupta, V.V., Jorquera, M.A.: Diversity, community structure, and potential functions of root- associated bacterial communities of different wheat (triticum aestivum) cultivars under field conditions. Agronomy13(5), 1392 (2023)

2023
[34]

Industrial Crops and Products244, 123231 (2026)

Wang, Y., Song, T., Liang, R., Zhang, Y., Wang, W., Cao, L., Li, L., Lin, Y.: Microbial inoculants alleviate continuous cropping obstacles in gynostemma pentaphyllum (thunb.) makino by reshaping rhizosphere microbial communities. Industrial Crops and Products244, 123231 (2026)

2026
[35]

Current opinion in microbiology44, 41–49 (2018)

Gonze, D., Coyte, K.Z., Lahti, L., Faust, K.: Microbial communities as dynamical systems. Current opinion in microbiology44, 41–49 (2018)

2018
[36]

Pediatric Research98(4), 1223–1231 (2025)

Neu, J., Stewart, C.J.: Neonatal microbiome in the multiomics era: development and its impact on long-term health. Pediatric Research98(4), 1223–1231 (2025)

2025
[37]

mSystems10(10), 00843–25 (2025)

Xu, Q., Zhang, Y., Tao, S.: Functional dynamics reveal the response of the crabap- ple (malus sp.) phyllosphere microbiome to gymnosporangium yamadae infection. mSystems10(10), 00843–25 (2025)

2025
[38]

Graph Attention Networks

Veliˇ ckovi´ c, P., Cucurull, G., Casanova, A., Romero, A., Lio, P., Bengio, Y.: Graph attention networks. arXiv preprint arXiv:1710.10903 (2017)

work page internal anchor Pith review Pith/arXiv arXiv 2017
[39]

Advances in neural information processing systems30(2017)

Hamilton, W., Ying, Z., Leskovec, J.: Inductive representation learning on large graphs. Advances in neural information processing systems30(2017)

2017
[40]

In: Wallach, H., Larochelle, H., Beygelzimer, A., Alch´ e-Buc, F., Fox, E., Gar- nett, R

Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., Antiga, L., Desmaison, A., Kopf, A., Yang, E., DeVito, Z., Raison, M., Tejani, A., Chilamkurthy, S., Steiner, B., Fang, L., Bai, J., Chin- tala, S.: Pytorch: An imperative style, high-performance deep learning library. In: Wallach, H., Larochelle, ...

2019

[1] [1]

Journal of Computational physics378, 686–707 (2019)

Raissi, M., Perdikaris, P., Karniadakis, G.E.: Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational physics378, 686–707 (2019)

2019

[2] [2]

arXiv preprint arXiv:2509.02282 (2025)

Fontanarrosa, P., Barnes, C.P.: Exploring accuracy and uncertainty quantification in physics-informed neural networks for inferring microbial community dynamics. arXiv preprint arXiv:2509.02282 (2025)

work page arXiv 2025

[3] [3]

bioRxiv (2026) https://doi.org/10.64898/2026.04.29.721522 https://www.biorxiv.org/content/early/2026/05/02/2026.04.29.721522.full.pdf

Gamage, C.J., Xia, Y., Rupasinghe, R., Senevirathne, S., Senanayake, D., Malepathirana, T., Hevapathige, A., Corbett, M., O’Brien, T.J., Petrou, S., Berkovic, S.F., Scheffer, I.E., Gecz, J., Bahlo, M., Bennett, M.F., Halgamuge, S.K.: Knowledge inclusive machine learning for disease gene prioritisation. bioRxiv (2026) https://doi.org/10.64898/2026.04.29.72...

work page doi:10.64898/2026.04.29.721522 2026

[4] [4]

American Society for Microbiology 1752 N St., NW, Washington, DC (2016)

Blaser, M.J., Cardon, Z.G., Cho, M.K., Dangl, J.L., Donohue, T.J., Green, J.L., Knight, R., Maxon, M.E., Northen, T.R., Pollard, K.S., et al.: Toward a predic- tive understanding of Earth’s microbiomes to address 21st century challenges. American Society for Microbiology 1752 N St., NW, Washington, DC (2016)

2016

[5] [5]

Nature metabolism1(1), 34–46 (2019)

Cani, P.D., Van Hul, M., Lefort, C., Depommier, C., Rastelli, M., Everard, A.: Microbial regulation of organismal energy homeostasis. Nature metabolism1(1), 34–46 (2019)

2019

[6] [6]

Cell148(6), 1258–1270 (2012)

Clemente, J.C., Ursell, L.K., Parfrey, L.W., Knight, R.: The impact of the gut microbiota on human health: an integrative view. Cell148(6), 1258–1270 (2012)

2012

[7] [7]

Advanced drug delivery reviews210, 115332 (2024) 29

Cho, Y.S., Han, K., Xu, J., Moon, J.J.: Novel strategies for modulating the gut microbiome for cancer therapy. Advanced drug delivery reviews210, 115332 (2024) 29

2024

[8] [8]

Bergey, D., Hendricks: Bergey’s Manual of Determinative Bacteriology, 9th edn. (1994)

1994

[9] [9]

Nature Reviews Microbiology10(8), 538–550 (2012)

Faust, K., Raes, J.: Microbial interactions: from networks to models. Nature Reviews Microbiology10(8), 538–550 (2012)

2012

[10] [10]

Elife11, 73870 (2022)

Baranwal, M., Clark, R.L., Thompson, J., Sun, Z., Hero, A.O., Venturelli, O.S.: Recurrent neural networks enable design of multifunctional synthetic human gut microbiome dynamics. Elife11, 73870 (2022)

2022

[11] [11]

BMC Molecular and Cell Biology21(Suppl 1), 34 (2020)

Vidanaarachchi, R., Shaw, M., Tang, S.-L., Halgamuge, S.: Imparo: inferring microbial interactions through parameter optimisation. BMC Molecular and Cell Biology21(Suppl 1), 34 (2020)

2020

[12] [12]

In: Forty-first International Conference on Machine Learning (2024)

Ruaud, A., Sancaktar, C., Bagatella, M., Ratzke, C., Martius, G.: Modelling microbial communities with graph neural networks. In: Forty-first International Conference on Machine Learning (2024)

2024

[13] [13]

Msystems8(3), 00757–22 (2023)

Dedrick, S., Warrier, V., Lemon, K.P., Momeni, B.: When does a lotka-volterra model represent microbial interactions? insights from in vitro nasal bacterial communities. Msystems8(3), 00757–22 (2023)

2023

[14] [14]

PLoS computational biology16(5), 1007917 (2020)

Joseph, T.A., Shenhav, L., Xavier, J.B., Halperin, E., Pe’er, I.: Compositional lotka-volterra describes microbial dynamics in the simplex. PLoS computational biology16(5), 1007917 (2020)

2020

[15] [15]

bioRxiv (2024)

Ma, C., Liu, S., Koslicki, D.: Metagenomickg: a knowledge graph for metagenomic applications. bioRxiv (2024)

2024

[16] [16]

Genome biology12(5), 50 (2011)

Caporaso, J.G., Lauber, C.L., Costello, E.K., Berg-Lyons, D., Gonzalez, A., Stombaugh, J., Knights, D., Gajer, P., Ravel, J., Fierer, N.,et al.: Moving pictures of the human microbiome. Genome biology12(5), 50 (2011)

2011

[17] [17]

Microbiology spectrum10(5), 01803–22 (2022)

Wang, C.-W., Michelle Wong, J.-W., Yeh, S.-S., Eric Hsieh, Y., Tseng, C.-H., Yang, S.-H., Tang, S.-L.: Soil bacterial community may offer solutions for ginger cultivation. Microbiology spectrum10(5), 01803–22 (2022)

2022

[18] [18]

Nature communications12(1), 3254 (2021)

Clark, R.L., Connors, B.M., Stevenson, D.M., Hromada, S.E., Hamilton, J.J., Amador-Noguez, D., Venturelli, O.S.: Design of synthetic human gut microbiome assembly and butyrate production. Nature communications12(1), 3254 (2021)

2021

[19] [19]

Microbiome6(1), 193 (2018)

Grier, A., McDavid, A., Wang, B., Qiu, X., Java, J., Bandyopadhyay, S., Yang, H., Holden-Wiltse, J., Kessler, H.A., Gill, A.L.,et al.: Neonatal gut and respiratory microbiota: coordinated development through time and space. Microbiome6(1), 193 (2018)

2018

[20] [20]

Microbiology Spectrum12(10), 04109–23 (2024)

Karwowska, Z., Szczerbiak, P., Kosciolek, T.: Microbiome time series data reveal 30 predictable patterns of change. Microbiology Spectrum12(10), 04109–23 (2024)

2024

[21] [21]

Nature Ecology & Evolution5(6), 747–756 (2021)

Averill, C., Werbin, Z.R., Atherton, K.F., Bhatnagar, J.M., Dietze, M.C.: Soil microbiome predictability increases with spatial and taxonomic scale. Nature Ecology & Evolution5(6), 747–756 (2021)

2021

[22] [22]

MBio13(1), 03161–21 (2022)

Armour, C.R., Top¸ cuo˘ glu, B.D., Garretto, A., Schloss, P.D.: A goldilocks princi- ple for the gut microbiome: taxonomic resolution matters for microbiome-based classification of colorectal cancer. MBio13(1), 03161–21 (2022)

2022

[23] [23]

Frontiers in Microbiology15, 1442208 (2024)

Cho, G., Kim, D.-R., Kwak, Y.-S.: Ecological shifts in soil microbiota and root rot disease progress during ginseng monoculture. Frontiers in Microbiology15, 1442208 (2024)

2024

[24] [24]

ISME communications1(1), 31 (2021)

Reitmeier, S., Hitch, T.C., Treichel, N., Fikas, N., Hausmann, B., Ramer-Tait, A.E., Neuhaus, K., Berry, D., Haller, D., Lagkouvardos, I.,et al.: Handling of spurious sequences affects the outcome of high-throughput 16s rrna gene amplicon profiling. ISME communications1(1), 31 (2021)

2021

[25] [25]

https://pubmed.ncbi.nlm.nih.gov/

National Library of Medicine: PubMed. https://pubmed.ncbi.nlm.nih.gov/. Accessed: 2025-09-29

2025

[26] [26]

Minilm: Deep self-attention distillation for task-agnostic compression of pre-trained transformers, 2020

Wang, X., Zhang, F.W., Wang, M., Li, J., Zhou, M.: Minilm: Deep self- attention distillation for task-agnostic compression of pretrained transformers. arXiv preprint arXiv:2002.10957 (2020)

work page arXiv 2002

[27] [27]

In: Proceedings of NAACL-HLT (2019)

Devlin, J., Chang, M.-W., Lee, K., Toutanova, K.: Bert: Pre-training of deep bidirectional transformers for language understanding. In: Proceedings of NAACL-HLT (2019)

2019

[28] [28]

bioRxiv (2024) https://doi.org/10.1101/2024.06.30.601456 https://www.biorxiv.org/content/early/2024/07/02/2024.06.30.601456.full.pdf

Vidanaarachchi, R., Tang, S.L., Halgamuge, S.: Copr: Collective pattern recognition—a framework for microbial community activ- ity analysis. bioRxiv (2024) https://doi.org/10.1101/2024.06.30.601456 https://www.biorxiv.org/content/early/2024/07/02/2024.06.30.601456.full.pdf

work page doi:10.1101/2024.06.30.601456 2024

[29] [29]

Anaerobe85, 102819 (2024)

Shin, J.H., Tillotson, G., MacKenzie, T.N., Warren, C.A., Wexler, H.M., Gold- stein, E.J.: Bacteroides and related species: The keystone taxa of the human gut microbiota. Anaerobe85, 102819 (2024)

2024

[30] [30]

Cell host & microbe31(4), 485–499 (2023)

Culp, E.J., Goodman, A.L.: Cross-feeding in the gut microbiome: ecology and mechanisms. Cell host & microbe31(4), 485–499 (2023)

2023

[31] [31]

MBio5(3), 10–1128 (2014)

Coyne, M.J., Zitomersky, N.L., McGuire, A.M., Earl, A.M., Comstock, L.E.: Evi- dence of extensive dna transfer between bacteroidales species within the human gut. MBio5(3), 10–1128 (2014)

2014

[32] [32]

Nature Biotechnology 41(10), 1424–1433 (2023)

Huang, Y., Sheth, R.U., Zhao, S., Cohen, L.A., Dabaghi, K., Moody, T., Sun, Y., 31 Ricaurte, D., Richardson, M., Velez-Cortes, F.,et al.: High-throughput micro- bial culturomics using automation and machine learning. Nature Biotechnology 41(10), 1424–1433 (2023)

2023

[33] [33]

Agronomy13(5), 1392 (2023)

Acu˜ na, J.J., Rilling, J.I., Inostroza, N.G., Manquian, J., Zhang, Q., Gupta, V.V., Jorquera, M.A.: Diversity, community structure, and potential functions of root- associated bacterial communities of different wheat (triticum aestivum) cultivars under field conditions. Agronomy13(5), 1392 (2023)

2023

[34] [34]

Industrial Crops and Products244, 123231 (2026)

Wang, Y., Song, T., Liang, R., Zhang, Y., Wang, W., Cao, L., Li, L., Lin, Y.: Microbial inoculants alleviate continuous cropping obstacles in gynostemma pentaphyllum (thunb.) makino by reshaping rhizosphere microbial communities. Industrial Crops and Products244, 123231 (2026)

2026

[35] [35]

Current opinion in microbiology44, 41–49 (2018)

Gonze, D., Coyte, K.Z., Lahti, L., Faust, K.: Microbial communities as dynamical systems. Current opinion in microbiology44, 41–49 (2018)

2018

[36] [36]

Pediatric Research98(4), 1223–1231 (2025)

Neu, J., Stewart, C.J.: Neonatal microbiome in the multiomics era: development and its impact on long-term health. Pediatric Research98(4), 1223–1231 (2025)

2025

[37] [37]

mSystems10(10), 00843–25 (2025)

Xu, Q., Zhang, Y., Tao, S.: Functional dynamics reveal the response of the crabap- ple (malus sp.) phyllosphere microbiome to gymnosporangium yamadae infection. mSystems10(10), 00843–25 (2025)

2025

[38] [38]

Graph Attention Networks

Veliˇ ckovi´ c, P., Cucurull, G., Casanova, A., Romero, A., Lio, P., Bengio, Y.: Graph attention networks. arXiv preprint arXiv:1710.10903 (2017)

work page internal anchor Pith review Pith/arXiv arXiv 2017

[39] [39]

Advances in neural information processing systems30(2017)

Hamilton, W., Ying, Z., Leskovec, J.: Inductive representation learning on large graphs. Advances in neural information processing systems30(2017)

2017

[40] [40]

In: Wallach, H., Larochelle, H., Beygelzimer, A., Alch´ e-Buc, F., Fox, E., Gar- nett, R

Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., Antiga, L., Desmaison, A., Kopf, A., Yang, E., DeVito, Z., Raison, M., Tejani, A., Chilamkurthy, S., Steiner, B., Fang, L., Bai, J., Chin- tala, S.: Pytorch: An imperative style, high-performance deep learning library. In: Wallach, H., Larochelle, ...

2019