Predicting protein condensate formation using machine learning

Mierlo, Guido van; Jansen, Jurriaan R.G.; Wang, Jie; Poser, Ina; Heeringen, Simon J. van; Vermeulen, Michiel

doi:10.1016/j.celrep.2021.108705

Cited by 85 publications

(103 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These differences indicate the importance of the liquid–liquid or liquid–solid phase separations of proteins. In addition, several methods have been developed to predict the PPS of proteins ( Vernon and Forman-Kay, 2019 ; Hardenberg et al, 2021 ; van Mierlo et al, 2021 ). These prediction methods can identify proteins with a high PPS from the proteome, such as the distribution of the amino acid residues and the complexity of the sequence.…”

Section: Discussionmentioning

confidence: 99%

Solubility Parameters of Amino Acids on Liquid–Liquid Phase Separation and Aggregation of Proteins

Nomoto

Nishinami

Shiraki

2021

Front. Cell Dev. Biol.

View full text Add to dashboard Cite

The solution properties of amino acids determine the folding, aggregation, and liquid–liquid phase separation (LLPS) behaviors of proteins. Various indices of amino acids, such as solubility, hydropathy, and conformational parameter, describe the behaviors of protein folding and solubility both in vitro and in vivo. However, understanding the propensity of LLPS and aggregation is difficult due to the multiple interactions among different amino acids. Here, the solubilities of aromatic amino acids (SAs) were investigated in solution containing 20 types of amino acids as amino acid solvents. The parameters of SAs in amino acid solvents (PSASs) were varied and dependent on the type of the solvent. Specifically, Tyr and Trp had the highest positive values while Glu and Asp had the lowest. The PSAS values represent soluble and insoluble interactions, which collectively are the driving force underlying the formation of droplets and aggregates. Interestingly, the PSAS of a soluble solvent reflected the affinity between amino acids and aromatic rings, while that of an insoluble solvent reflected the affinity between amino acids and water. These findings suggest that the PSAS can distinguish amino acids that contribute to droplet and aggregate formation, and provide a deeper understanding of LLPS and aggregation of proteins.

show abstract

Section: Discussionmentioning

confidence: 99%

Solubility Parameters of Amino Acids on Liquid–Liquid Phase Separation and Aggregation of Proteins

Nomoto

Nishinami

Shiraki

2021

Front. Cell Dev. Biol.

View full text Add to dashboard Cite

show abstract

“…While both studies identified hundreds of candidate R-loop binding proteins (RLBPs) they overlapped by >100 proteins creating a higher confidence set of proteins that may preferentially recognize R-loops. Building on this set of high-confidence RLBPs it is possible to extract common protein features at the amino acid sequence level, and at a higher level such as protein domain organization or function (10)(11)(12). Based on this, here we use a random forest machine learning classifier to score R-loop binding potential for the human proteome.…”

Section: Introductionmentioning

confidence: 99%

Global prediction of candidate R-loop binding and R-loop regulatory proteins

Fournier

Kumar

Smith

et al. 2021

Preprint

View full text Add to dashboard Cite

In the past decade there has been a growing appreciation for R-loop structures as important regulators of the epigenome, telomere maintenance, DNA repair and replication. Given these numerous functions, dozens, or potentially hundreds, of proteins could serve as direct or indirect regulators of R-loop writing, reading, and erasing. In order to understand common properties shared amongst potential R-loop binding proteins (RLBPs) we mined published proteomic studies and distilled 10 features that were enriched in RLBPs compared to the rest of the proteome. We used these RLBP-specific features along with their amino acid composition to create a random forest classifier which predicts the likelihood of a protein to bind to R-loops. In parallel, we employed a whole-genome CRISPR screen coupled with flow-cytometry using the S9.6 monoclonal antibody to sort guide RNAs associated with induction of high S9.6 staining. Known R-loop regulating pathways such as splicing and DNA damage repair are highly enriched in our datasets, and we validate two new R-loop modulating proteins. Together these resources provide a reference to pursue analyses of novel R-loop regulatory proteins.

show abstract

“…We also analyzed PML isoforms for susceptibility to phase separation using the PSPredictor software package [ 53 ] based on machine learning algorithms. This analysis showed that the amino acid sequences of all the studied isoforms are characteristic of proteins capable of LLPS.…”

Section: Resultsmentioning

confidence: 99%

“…This tool enables the rapid generation of disorder profile plots for individual polypeptides as well as arrays of polypeptides. The prediction of LLPS PML-isoforms was performed using by PSPredictor software package [ 53 ].…”

Section: Methodsmentioning

confidence: 99%

The Role of Non-Specific Interactions in Canonical and ALT-Associated PML-Bodies Formation and Dynamics

Fonin

Silonov

Shpironok

et al. 2021

IJMS

View full text Add to dashboard Cite

In this work, we put forward a hypothesis about the decisive role of multivalent nonspecific interactions in the early stages of PML body formation. Our analysis of the PML isoform sequences showed that some of the PML isoforms, primarily PML-II, are prone to phase separation due to their polyampholytic properties and the disordered structure of their C-terminal domains. The similarity of the charge properties of the C-terminal domains of PML-II and PML-VI isoforms made it possible for the first time to detect migration of PML-VI from PML bodies to the periphery of the cell nucleus, similar to the migration of PML-II isoforms. We found a population of “small” (area less than 1 µm2) spherical PML bodies with high dynamics of PML isoforms exchange with nucleoplasm and a low fraction of immobilized proteins, which indicates their liquid state properties. Such structures can act as “seeds” of functionally active PML bodies, providing the necessary concentration of PML isoforms for the formation of intermolecular disulfide bonds between PML monomers. FRAP analysis of larger bodies of toroidal topology showed the existence of an insoluble scaffold in their structure. The hypothesis about the role of nonspecific multiple weak interactions in the formation of PML bodies is further supported by the change in the composition of the scaffold proteins of PML bodies, but not their solidification, under conditions of induction of dimerization of PML isoforms under oxidative stress. Using the colocalization of ALT-associated PML bodies (APBs) with TRF1, we identified APBs and showed the difference in the dynamic properties of APBs and canonical PML bodies.

show abstract

Predicting protein condensate formation using machine learning

Abstract: Highlights d An overview of amino acid features associated with condensate-forming (PPS) proteins d Development of a machine learning classifier (PSAP) to predict candidate PPS proteins d PSAP enabled the discovery of new PPS proteins, including DAZAP1 and CPEB3

Cited by 85 publications

References 60 publications

Solubility Parameters of Amino Acids on Liquid–Liquid Phase Separation and Aggregation of Proteins

Solubility Parameters of Amino Acids on Liquid–Liquid Phase Separation and Aggregation of Proteins

Global prediction of candidate R-loop binding and R-loop regulatory proteins

The Role of Non-Specific Interactions in Canonical and ALT-Associated PML-Bodies Formation and Dynamics

Contact Info

Product

Resources

About