1
|
Das S, Ramanan N, Kunapuli G, Radivojac P, Natarajan S. Active feature elicitation: An unified framework. Front Artif Intell 2023; 6:1029943. [PMID: 37035530 PMCID: PMC10080103 DOI: 10.3389/frai.2023.1029943] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/28/2022] [Accepted: 03/03/2023] [Indexed: 04/11/2023] Open
Abstract
We consider the problem of active feature elicitation in which, given some examples with all the features (say, the full Electronic Health Record), and many examples with some of the features (say, demographics), the goal is to identify the set of examples on which more information (say, lab tests) need to be collected. The observation is that some set of features may be more expensive, personal or cumbersome to collect. We propose a classifier-independent, similarity metric-independent, general active learning approach which identifies examples that are dissimilar to the ones with the full set of data and acquire the complete set of features for these examples. Motivated by four real clinical tasks, our extensive evaluation demonstrates the effectiveness of this approach. To demonstrate the generalization capabilities of the proposed approach, we consider different divergence metrics and classifiers and present consistent results across the domains.
Collapse
Affiliation(s)
- Srijita Das
- Department of Computing Science, University of Alberta, Edmonton, AB, Canada
| | | | - Gautam Kunapuli
- Computer Science Department, University of Texas at Dallas, Dallas, TX, United States
| | - Predrag Radivojac
- Khoury College of Computer Sciences, Northeastern University, Boston, MA, United States
| | - Sriraam Natarajan
- Computer Science Department, University of Texas at Dallas, Dallas, TX, United States
| |
Collapse
|
2
|
Karunakaran KB, Yanamala N, Boyce G, Becich MJ, Ganapathiraju MK. Malignant Pleural Mesothelioma Interactome with 364 Novel Protein-Protein Interactions. Cancers (Basel) 2021; 13:1660. [PMID: 33916178 PMCID: PMC8037232 DOI: 10.3390/cancers13071660] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/28/2021] [Revised: 03/19/2021] [Accepted: 03/22/2021] [Indexed: 12/20/2022] Open
Abstract
Malignant pleural mesothelioma (MPM) is an aggressive cancer affecting the outer lining of the lung, with a median survival of less than one year. We constructed an 'MPM interactome' with over 300 computationally predicted protein-protein interactions (PPIs) and over 2400 known PPIs of 62 literature-curated genes whose activity affects MPM. Known PPIs of the 62 MPM associated genes were derived from Biological General Repository for Interaction Datasets (BioGRID) and Human Protein Reference Database (HPRD). Novel PPIs were predicted by applying the HiPPIP algorithm, which computes features of protein pairs such as cellular localization, molecular function, biological process membership, genomic location of the gene, and gene expression in microarray experiments, and classifies the pairwise features as interacting or non-interacting based on a random forest model. We validated five novel predicted PPIs experimentally. The interactome is significantly enriched with genes differentially ex-pressed in MPM tumors compared with normal pleura and with other thoracic tumors, genes whose high expression has been correlated with unfavorable prognosis in lung cancer, genes differentially expressed on crocidolite exposure, and exosome-derived proteins identified from malignant mesothelioma cell lines. 28 of the interactors of MPM proteins are targets of 147 U.S. Food and Drug Administration (FDA)-approved drugs. By comparing disease-associated versus drug-induced differential expression profiles, we identified five potentially repurposable drugs, namely cabazitaxel, primaquine, pyrimethamine, trimethoprim and gliclazide. Preclinical studies may be con-ducted in vitro to validate these computational results. Interactome analysis of disease-associated genes is a powerful approach with high translational impact. It shows how MPM-associated genes identified by various high throughput studies are functionally linked, leading to clinically translatable results such as repurposed drugs. The PPIs are made available on a webserver with interactive user interface, visualization and advanced search capabilities.
Collapse
Affiliation(s)
- Kalyani B. Karunakaran
- Supercomputer Education and Research Centre, Indian Institute of Science, Bangalore 560012, India;
| | - Naveena Yanamala
- Exposure Assessment Branch, National Institute of Occupational Safety and Health, Center for Disease Control, Morgantown, WV 26506, USA; (N.Y.); (G.B.)
| | - Gregory Boyce
- Exposure Assessment Branch, National Institute of Occupational Safety and Health, Center for Disease Control, Morgantown, WV 26506, USA; (N.Y.); (G.B.)
| | - Michael J. Becich
- Department of Biomedical Informatics, School of Medicine, University of Pittsburgh, Pittsburgh, PA 15206, USA;
| | - Madhavi K. Ganapathiraju
- Department of Biomedical Informatics, School of Medicine, University of Pittsburgh, Pittsburgh, PA 15206, USA;
- Intelligent Systems Program, School of Computing and Information, University of Pittsburgh, Pittsburgh, PA 15213, USA
| |
Collapse
|
3
|
Balkenhol J, Kaltdorf KV, Mammadova-Bach E, Braun A, Nieswandt B, Dittrich M, Dandekar T. Comparison of the central human and mouse platelet signaling cascade by systems biological analysis. BMC Genomics 2020; 21:897. [PMID: 33353544 PMCID: PMC7756956 DOI: 10.1186/s12864-020-07215-4] [Citation(s) in RCA: 9] [Impact Index Per Article: 2.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/09/2020] [Accepted: 11/08/2020] [Indexed: 12/12/2022] Open
Abstract
Background Understanding the molecular mechanisms of platelet activation and aggregation is of high interest for basic and clinical hemostasis and thrombosis research. The central platelet protein interaction network is involved in major responses to exogenous factors. This is defined by systemsbiological pathway analysis as the central regulating signaling cascade of platelets (CC). Results The CC is systematically compared here between mouse and human and major differences were found. Genetic differences were analysed comparing orthologous human and mouse genes. We next analyzed different expression levels of mRNAs. Considering 4 mouse and 7 human high-quality proteome data sets, we identified then those major mRNA expression differences (81%) which were supported by proteome data. CC is conserved regarding genetic completeness, but we observed major differences in mRNA and protein levels between both species. Looking at central interactors, human PLCB2, MMP9, BDNF, ITPR3 and SLC25A6 (always Entrez notation) show absence in all murine datasets. CC interactors GNG12, PRKCE and ADCY9 occur only in mice. Looking at the common proteins, TLN1, CALM3, PRKCB, APP, SOD2 and TIMP1 are higher abundant in human, whereas RASGRP2, ITGB2, MYL9, EIF4EBP1, ADAM17, ARRB2, CD9 and ZYX are higher abundant in mouse. Pivotal kinase SRC shows different regulation on mRNA and protein level as well as ADP receptor P2RY12. Conclusions Our results highlight species-specific differences in platelet signaling and points of specific fine-tuning in human platelets as well as murine-specific signaling differences. Supplementary Information The online version contains supplementary material available at 10.1186/s12864-020-07215-4.
Collapse
Affiliation(s)
- Johannes Balkenhol
- Functional Genomics and Systems Biology Group, Department of Bioinformatics, Biocenter, Am Hubland, University of Würzburg, D-97074, Würzburg, Germany
| | - Kristin V Kaltdorf
- Functional Genomics and Systems Biology Group, Department of Bioinformatics, Biocenter, Am Hubland, University of Würzburg, D-97074, Würzburg, Germany
| | - Elmina Mammadova-Bach
- Institute of Experimental Biomedicine, University Hospital and Rudolf Virchow Centre, University of Würzburg, Würzburg, Germany.,Present address: Division of Nephrology, Department of Medicine IV, Hospital of the Ludwig, Maximilian University of Munich, D-80336, Munich, Germany
| | - Attila Braun
- Member of the German Center for Lung Research (DZL), Walther-Straub-Institute for Pharmacology and Toxicology, Ludwig-Maximilians University Munich, Munich, Germany
| | - Bernhard Nieswandt
- Institute of Experimental Biomedicine, University Hospital and Rudolf Virchow Centre, University of Würzburg, Würzburg, Germany
| | - Marcus Dittrich
- Functional Genomics and Systems Biology Group, Department of Bioinformatics, Biocenter, Am Hubland, University of Würzburg, D-97074, Würzburg, Germany.,Dept of Genetics, Biocenter, Am Hubland, University of Würzburg, Am Hubland, D 97074, Würzburg, Germany
| | - Thomas Dandekar
- Functional Genomics and Systems Biology Group, Department of Bioinformatics, Biocenter, Am Hubland, University of Würzburg, D-97074, Würzburg, Germany.
| |
Collapse
|
4
|
Schizophrenia interactome with 504 novel protein-protein interactions. NPJ SCHIZOPHRENIA 2016; 2:16012. [PMID: 27336055 PMCID: PMC4898894 DOI: 10.1038/npjschz.2016.12] [Citation(s) in RCA: 39] [Impact Index Per Article: 4.9] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 10/21/2015] [Revised: 02/17/2016] [Accepted: 02/23/2016] [Indexed: 11/29/2022]
Abstract
Genome-wide association studies of schizophrenia (GWAS) have revealed the role of rare and common genetic variants, but the functional effects of the risk variants remain to be understood. Protein interactome-based studies can facilitate the study of molecular mechanisms by which the risk genes relate to schizophrenia (SZ) genesis, but protein–protein interactions (PPIs) are unknown for many of the liability genes. We developed a computational model to discover PPIs, which is found to be highly accurate according to computational evaluations and experimental validations of selected PPIs. We present here, 365 novel PPIs of liability genes identified by the SZ Working Group of the Psychiatric Genomics Consortium (PGC). Seventeen genes that had no previously known interactions have 57 novel interactions by our method. Among the new interactors are 19 drug targets that are targeted by 130 drugs. In addition, we computed 147 novel PPIs of 25 candidate genes investigated in the pre-GWAS era. While there is little overlap between the GWAS genes and the pre-GWAS genes, the interactomes reveal that they largely belong to the same pathways, thus reconciling the apparent disparities between the GWAS and prior gene association studies. The interactome including 504 novel PPIs overall, could motivate other systems biology studies and trials with repurposed drugs. The PPIs are made available on a webserver, called Schizo-Pi at http://severus.dbmi.pitt.edu/schizo-pi with advanced search capabilities.
Collapse
|