Individual-specific networks for prediction modelling – A scoping review of methods

Abstract

Background Recent advances in biotechnology enable the acquisition of high-dimensional data on individuals, posing challenges for prediction models which traditionally use covariates such as clinical patient characteristics. Alternative forms of covariate representations for the features derived from these modern data modalities should be considered that can utilize their intrinsic interconnection. The connectivity information between these features can be represented as an individual-specific network defined by a set of nodes and edges, the strength of which can vary from individual to individual. Global or local graph-theoretical features describing the network may constitute potential prognostic biomarkers instead of or in addition to traditional covariates and may replace the often unsuccessful search for individual biomarkers in a high-dimensional predictor space. Methods We conducted a scoping review to identify, collate and critically appraise the state-of-art in the use of individual-specific networks for prediction modelling in medicine and applied health research, published during 2000–2020 in the electronic databases PubMed, Scopus and Embase. Results Our scoping review revealed the main application areas namely neurology and pathopsychology, followed by cancer research, cardiology and pathology (N = 148). Network construction was mainly based on Pearson correlation coefficients of repeated measurements, but also alternative approaches (e.g. partial correlation, visibility graphs) were found. For covariates measured only once per individual, network construction was mostly based on quantifying an individual’s contribution to the overall group-level structure. Despite the multitude of identified methodological approaches for individual-specific network inference, the number of studies that were intended to enable the prediction of clinical outcomes for future individuals was quite limited, and most of the models served as proof of concept that network characteristics can in principle be useful for prediction. Conclusion The current body of research clearly demonstrates the value of individual-specific network analysis for prediction modelling, but it has not yet been considered as a general tool outside the current areas of application. More methodological research is still needed on well-founded strategies for network inference, especially on adequate network sparsification and outcome-guided graph-theoretical feature extraction and selection, and on how networks can be exploited efficiently for prediction modelling.

Flexible parametrization of graph‐theoretical features from individual‐specific networks for prediction

“…Many neuroimaging studies have demonstrated the potential of functional network connectivity patterns estimated from rs-fMRI to discriminate groups or to predict a clinical outcome. 17 Two modeling approaches are mainly employed in applied studies to address the variability of graph-theoretical features resulting from the selection of thresholds across a broad spectrum of sparsity levels. The first one selects a single threshold based on optimal prediction (OPT), while the other one averages the resulting graph-theoretical feature over a predefined range of thresholds (AVG).…”

“…Commonly, it is either a maximum value for the density (density-based thresholding) or a minimum edge weight (weight-based thresholding). 17 Approaches to sparsify the individual-specific networks are not tied to the specific method used to estimate the connection strength. Neither for density-based nor for weight-based thresholding is there a commonly accepted consensus on the exact threshold that should be used to infer the individual-specific networks A i , and various methods are used in practice to determine the threshold at which graph-theoretical characteristics become relevant for prediction.…”

“…In case of a correlation-derived adjacency matrix, the search is commonly performed in T = {0.1, … , 0.4} with a step size of 0.05 or 0.1 and the (apparent) mean squared prediction error as the loss function to be minimized. 17…”

Edge and modular significance assessment in individual-specific networks

“…Common practice is to aggregate information, such as averaging edge weights in each ISN, and then look for associations with phenotypes of interest (for instance, drug reaction and time-to-clinical-event 22 , 23 ). The most common objective of studies that include ISNs as input is prediction (for a review, see 24 ). This usually involves extracting graph-theoretical features and linking them to a phenotype of interest.…”

netANOVA: novel graph clustering technique with significance assessment via hierarchical ANOVA

“…Most of the graph analyses for complex diseases aggregate information across a whole cohort, failing to detect individual characteristics [ 5 ]. Exploiting individual-specific interactions rather than population-level systems will capture the heterogeneity between individuals and enhance the identification of new biomarkers for precision medicine.…”