Biology subjects

Cavallo, L.

Publications and source records attributed to Cavallo, L..

4 recordsLinked to original sources

bioRxiv2022

Improving classification of correct and incorrect protein-protein docking models by augmenting the training set

Protein-protein interactions drive many relevant biological events, such as infection, replication, and recognition. To control or engineer such events, we need to access the molecular details of the interaction provided by experimental 3D structures. However, such experiments take time and are expensive; moreover, the current technology cannot keep up with the high discovery rate of new interactions. Computational modeling, like protein-protein docking, can help to fill this gap by generating docking poses. Protein-protein docking generally consists of two parts, sampling and scoring. The sampling is an exhaustive search of the tridimensional space. The caveat of the sampling produces a large number of incorrect poses, producing a highly unbalanced dataset. This limits the utility of the data to train machine learning classifiers. Using weak supervision, we developed a data augmentation method that we named hAIkal. Using hAIkal, we increased the labeled training data to train several algorithms. We trained and obtained different classifiers; the best classifier has 81% accuracy and 0.51 MCC on the test set, surpassing the state-of-the-art scoring functions.

bioinformatics↗

bioRxiv2022

Duplicated zebrafish (Danio rerio) inositol phosphatases inpp5ka and inpp5kb diverged in expression pattern and function

One hurdle in the development of zebrafish models of human disease is the presence of multiple zebrafish orthologs resulting from whole genome duplication in teleosts. Mutations in Inositol polyphosphate 5-phosphatase K (INPP5K) lead to a syndrome characterized by variable presentation of intellectual disability, brain abnormalities, cataracts, muscle disease, and short stature. INPP5K is a phosphatase acting at position 5 of phosphoinositides to control their homeostasis and is involved in insulin signaling, cytoskeletal regulation, and protein trafficking. Previously, our group and others have replicated the human phenotypes in zebrafish knockdown models by targeting both INPP5K orthologs inpp5ka and inpp5kb. Here, we show that inpp5ka is the more closely related orthologue to human INPP5K. While both inpp5ka and inpp5kb mRNA expression levels follow a similar trend in the developing head, eyes, and tail, inpp5ka is much more abundantly expressed in these tissues than inpp5kb. In situ hybridization revealed a similar trend, also showing unique localization of inpp5kb in the pineal gland indicating different transcriptional regulation. We also found that inpp5kb has lost its catalytic activity against its preferred substrate, PtdIns(4,5)P2. Since most human mutations are missense changes disrupting phosphatase activity, we propose that loss of inpp5ka alone can be targeted to recapitulate the human presentation. In addition, we show that the function of inpp5kb has diverged from inpp5ka and may play a novel role in the zebrafish.

developmental biology↗

bioRxiv2021

A Random Forest Classifier for Protein-Protein Docking Models

Herein, we present the results of a machine learning approach we developed to single out correct 3D docking models of protein-protein complexes obtained by popular docking software. To this aim, we generated a set of {approx}7x106 docking models with three different docking programs (HADDOCK, FTDock and ZDOCK) for the 230 complexes in the protein-protein interaction benchmark, version 5 (BM5). Three different machine-learning approaches (Random Forest, Supported Vector Machine and Perceptron) were used to train classifiers with 158 different scoring functions (features). The Random Forest algorithm outperformed the other two algorithms and was selected for further optimization. Using a features selection algorithm, and optimizing the random forest hyperparameters, allowed us to train and validate a random forest classifier, named CoDES (COnservation Driven Expert System). Testing of CoDES on independent datasets, as well as results of its comparative performance with machine-learning methods recently developed in the field for the scoring of docking decoys, confirm its state-of-the-art ability to discriminate correct from incorrect decoys both in terms of global parameters and in terms of decoys ranked at the top positions.

bioinformatics↗

bioRxiv2020

D936Y and Other Mutations in the Fusion Core of the SARS-Cov-2 Spike Protein Heptad Repeat 1 Undermine the Post-Fusion Assembly

The iconic "red crown" of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is made of its spike (S) glycoprotein. The S protein is the Trojan horse of coronaviruses, mediating their entry into the host cells. While SARS-CoV-2 was becoming a global threat, scientists have been accumulating data on the virus at an impressive pace, both in terms of genomic sequences and of three-dimensional structures. On April 21st, the GISAID resource had collected 10,823 SARS-CoV-2 genomic sequences. We extracted from them all the complete S protein sequences and identified point mutations thereof. Six mutations were located on a 14-residue segment (929-943) in the "fusion core" of the heptad repeat 1 (HR1). Our modeling in the pre- and post-fusion S protein conformations revealed, for three of them, the loss of interactions stabilizing the post-fusion assembly. On May 29th, the SARS-CoV-2 genomic sequences in GISAID were 34,805. An analysis of the occurrences of the HR1 mutations in this updated dataset revealed a significant increase for the S929I and S939F mutations and a dramatic increase for the D936Y mutation, which was particularly widespread in Sweden and Wales/England. We notice that this is also the mutation causing the loss of a strong inter-monomer interaction, the D936-R1185 salt bridge, thus clearly weakening the post-fusion assembly.

bioinformatics↗