PairGP: Gaussian process modeling of longitudinal data from paired multi-condition studies

Vantini, Michele; Mannerström, Henrik; Rautio, Sini; Ahlfors, Helena; Stockinger, Brigitta; Lähdesmäki, Harri

doi:10.1016/j.compbiomed.2022.105268

Cited by 2 publications

(1 citation statement)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Especially, this allows to tackle complex experimental setups by thinking in terms of similarity matrices and curve smoothness (correlation matrix). Another related model is PairGP [37], likewise applied on gene expression time series and introducing GP regression to deal with multiple conditions and replicates. A main difference lies in the statistical testing framework suggested by the authors.…”

Section: Discussionmentioning

confidence: 99%

Hierarchical Gaussian process models explore the dark meltome of thermal proteome profiling experiments

Le Sueur,

Rattray,

Savitski

2023

Preprint

View full text Add to dashboard Cite

Thermal proteome profiling (TPP) is a proteome wide technology that enables unbiased detection of protein drug interactions as well as changes in post-translational state of proteins between different biological conditions. Statistical analysis of temperature range TPP (TPP-TR) datasets relies on comparing protein melting curves, describing the amount of non-denatured proteins as a function of temperature, between different conditions (e.g. presence or absence of a drug). However, state-of-the-art models are restricted to sigmoidal melting behaviours while unconventional melting curves, representing up to 50% of TPP-TR datasets, have recently been shown to carry important biological information.We present a novel statistical framework, based on hierarchical Gaussian process models and named GPMelt, to make TPP-TR datasets analysis unbiased with respect to the melting profiles of proteins. GPMelt scales to multiple conditions, and extension of the model to deeper hierarchies (i.e. with additional sub-levels) allows to deal with complex TPP-TR protocols. Collectively, our statistical framework extends the analysis of TPP-TR datasets for both protein and peptide level melting curves, offering access to thousands of previously excluded melting curves and thus substantially increasing the coverage and the ability of TPP to uncover new biology.Author summaryProteins interactions with other proteins, nucleic acids or metabolites, are key to all biological processes. Being able to detect these interactions is essential to understand biological systems. Thermal proteome profiling is a proteome-wide biological assay able to capture these interactions. It consists in analysing the effect of heat treatment on proteins. Indeed, proteins, under physiological conditions, are folded. This folding gets disrupted as the temperature increases. The way this unfolding happens, called the melting profile of the protein, informs on the interactions of proteins. For example, the interaction of a protein with another protein can increase (thermally stabilise) or decrease (thermally destabilise) the temperature at which this protein starts unfolding. In this work, we present a new statistical method, named GPMelt, to analyse these melting profiles. Notably, GPMelt allows to analyse any melting profiles, independently of their shapes. The proposed improvements over previously published methods allow to investigate more robustly the melting profiles of more proteins, hence increasing the ability of thermal proteome profiling assays to discover new protein interactions. We anticipate that these advancements will aid in unravelling complex biological phenomena.

show abstract

Section: Discussionmentioning

confidence: 99%