Using neural networks to bridge scales in cancer: Mapping signaling pathways to phenotypes

Kim, Eun-Jung; Gerlee, Philip; Anderson, Alexander R.A.

doi:10.1101/324038

Cited by 2 publications

(5 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Only a few studies have employed ANN in the study of signaling pathways (43, 44). For example, (44) developed an ANN of 96 genes with A 3-layered Multi-perceptron with backpropagation learning and sigmoid activation function for the study of biomarkers in children sarcomas (44).…”

Section: Discussionmentioning

confidence: 99%

“…For example, (44) developed an ANN of 96 genes with A 3-layered Multi-perceptron with backpropagation learning and sigmoid activation function for the study of biomarkers in children sarcomas (44). Moreover (43) used a simplified neural network to integrate some of the environmental and molecular characteristics of cancer progression. This network was comprised by two microenvironmental input nodes (growth factor and death signal), and two phenotype output nodes (pro-growth and pro-death, (43).…”

Section: Discussionmentioning

confidence: 99%

“…Moreover (43) used a simplified neural network to integrate some of the environmental and molecular characteristics of cancer progression. This network was comprised by two microenvironmental input nodes (growth factor and death signal), and two phenotype output nodes (pro-growth and pro-death, (43). However, to our knowledge, this is the first study that combines structural data, docking simulations and ANN in the modulation of a SP.…”

Section: Discussionmentioning

confidence: 99%

“…In this aspect, the development of holistic computational methodologies of SP and their modulatory mechanisms can lead to the discovery of new pharmacological targets in complex disorders such as PD (41, 42). In this aspect, the use of ANN (artificial neural networks) have been previously used in a simplified model of the Toll-like receptor signaling pathway, and the PI3K/AKT pathway in the context of cancer (43).…”

Section: Introductionmentioning

confidence: 99%

“…Through the integration of structural and systemic modeling with artificial neural networks (ANN) approaches, it was possible to determine the potential of the nicotine analogs as neuroprotective agents and improve drug scanning efficiency. Even though ANN has been used to study gene interactions (44) or signaling in cancer (43), we use ANN as part of a novel computational strategy aiming to predict novel neuroprotective therapeutic compounds. We present a diagrammatic overview of the interaction algorithm summarizing the presented strategy allowing reproducibility and robustness of the approach (Fig 1).…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Novel computational deep learning strategy for neuroprotection identification reveals unique set of nicotine analogs as potential therapeutic compounds against Parkinson’s disease

Rojas-Rodríguez

Morantes

Pinzón

et al. 2019

Preprint

View full text Add to dashboard Cite

Dopaminergic replacement has been used for Parkinson’s Disease (PD) treatment with positive effects on motor symptomatology but with low effects over disease progression and prevention. Different epidemiological studies have shown that nicotine consumption decreases PD prevalence through the activation of neuroprotective mechanisms. Nicotine-induced neuroprotection has been associated with the overstimulation of intracellular signaling pathways (SP) such as Phosphatidyl Inositol 3-kinase/Protein kinase-B (PI3K/AKT) through nicotinic acetylcholine receptors (e.g α7 nAChRs) and the over-expression of the anti-apoptotic gene Bcl-2. Considering its harmful effects (toxicity and dependency), the search for nicotine analogs with decreased secondary effects, but similar neuroprotective activity, remains a promissory field of study. In this work, a computational strategy integrating structural bioinformatics, signaling pathway (SP) manual reconstruction, and deep learning was performed to predict the potential neuroprotective activity of a series of 8 novel nicotine analogs over the behavior of PI3K/AKT. We performed a protein-ligand analysis between nicotine analogs and α7 nAChRs receptor using geometrical conformers, physicochemical characterization of the analogs and developed a manually curated neuroprotective dataset to analyze their potential activity. Additionally, we developed a predictive machine-learning model for neuroprotection in PD through the integration of Markov Chain Monte-Carlo transition matrix for the SP with synthetic training datasets of the physicochemical properties and structural dataset. Our model was able to predict the potential neuroprotective activity of seven new nicotine analogs based on the binomial Bcl-2 response regulated by the activation of PI3K/AKT. We present a new computational strategy to predict the pharmacological neuroprotective potential of nicotine analogs based on SP architecture, using deep learning and structural data. Our theoretical strategy can be further applied to the study new treatments related with SP deregulation and may ultimately offer new opportunities for therapeutic interventions in neurodegenerative diseases.Author SummaryParkinson’s disease is one of the most prevalent neurodegenerative diseases across population over age 50. Affecting controlled movements and non-motor symptoms, treatments for Parkinson prevention are indispensable to reduce patient’s population in the future. Epidemiological data provide evidence that nicotine have a neuroprotective effect decreasing Parkinson prevalence. By interacting with nicotine receptors in neurons and modulating signaling pathways expressing anti-apoptotic genes nicotine arise as a putative neuroprotective therapy. Nevertheless, toxicity and dependency prevent the use of nicotine as a suitable drug. Nicotine analogs, structurally similar compounds emerge as an alternative for Parkinson preventive treatment. In this sense we developed a quantitative strategy to predict the potential neuroprotective activity of nicotine analogs. Our model is the first approach to predict neuroprotection in the context of Parkinson and signaling pathways using machine learning and computational chemistry.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Novel computational deep learning strategy for neuroprotection identification reveals unique set of nicotine analogs as potential therapeutic compounds against Parkinson’s disease

Rojas-Rodríguez

Morantes

Pinzón

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

Knowledge-primed neural networks enable biologically interpretable deep learning on single-cell sequencing data

Fortelny

Bock

2020

Genome Biol

View full text Add to dashboard Cite

Background Deep learning has emerged as a versatile approach for predicting complex biological phenomena. However, its utility for biological discovery has so far been limited, given that generic deep neural networks provide little insight into the biological mechanisms that underlie a successful prediction. Here we demonstrate deep learning on biological networks, where every node has a molecular equivalent, such as a protein or gene, and every edge has a mechanistic interpretation, such as a regulatory interaction along a signaling pathway. Results With knowledge-primed neural networks (KPNNs), we exploit the ability of deep learning algorithms to assign meaningful weights in multi-layered networks, resulting in a widely applicable approach for interpretable deep learning. We present a learning method that enhances the interpretability of trained KPNNs by stabilizing node weights in the presence of redundancy, enhancing the quantitative interpretability of node weights, and controlling for uneven connectivity in biological networks. We validate KPNNs on simulated data with known ground truth and demonstrate their practical use and utility in five biological applications with single-cell RNA-seq data for cancer and immune cells. Conclusions We introduce KPNNs as a method that combines the predictive power of deep learning with the interpretability of biological networks. While demonstrated here on single-cell sequencing data, this method is broadly relevant to other research areas where prior domain knowledge can be represented as networks.

show abstract

Using neural networks to bridge scales in cancer: Mapping signaling pathways to phenotypes

Cited by 2 publications

References 47 publications

Novel computational deep learning strategy for neuroprotection identification reveals unique set of nicotine analogs as potential therapeutic compounds against Parkinson’s disease

Novel computational deep learning strategy for neuroprotection identification reveals unique set of nicotine analogs as potential therapeutic compounds against Parkinson’s disease

Knowledge-primed neural networks enable biologically interpretable deep learning on single-cell sequencing data

Contact Info

Product

Resources

About