Journal of Personalized Medicine

Research

Jump to: Review

26 pages, 4805 KiB

Open AccessArticle

Dopaminergic Gene Dosage Reveals Distinct Biological Partitions between Autism and Developmental Delay as Revealed by Complex Network Analysis and Machine Learning Approaches

by André Santos, Francisco Caramelo, Joana Barbosa Melo and Miguel Castelo-Branco

J. Pers. Med. 2022, 12(10), 1579; https://doi.org/10.3390/jpm12101579 - 25 Sep 2022

Cited by 2 | Viewed by 2175

Abstract

The neurobiological mechanisms underlying Autism Spectrum Disorders (ASD) remains controversial. One factor contributing to this debate is the phenotypic heterogeneity observed in ASD, which suggests that multiple system disruptions may contribute to diverse patterns of impairment which have been reported between and within [...] Read more.

The neurobiological mechanisms underlying Autism Spectrum Disorders (ASD) remains controversial. One factor contributing to this debate is the phenotypic heterogeneity observed in ASD, which suggests that multiple system disruptions may contribute to diverse patterns of impairment which have been reported between and within study samples. Here, we used SFARI data to address genetic imbalances affecting the dopaminergic system. Using complex network analysis, we investigated the relations between phenotypic profiles, gene dosage and gene ontology (GO) terms related to dopaminergic neurotransmission from a polygenic point-of-view. We observed that the degree of distribution of the networks matched a power-law distribution characterized by the presence of hubs, gene or GO nodes with a large number of interactions. Furthermore, we identified interesting patterns related to subnetworks of genes and GO terms, which suggested applicability to separation of clinical clusters (Developmental Delay (DD) versus ASD). This has the potential to improve our understanding of genetic variability issues and has implications for diagnostic categorization. In ASD, we identified the separability of four key dopaminergic mechanisms disrupted with regard to receptor binding, synaptic physiology and neural differentiation, each belonging to particular subgroups of ASD participants, whereas in DD a more unitary biological pattern was found. Finally, network analysis was fed into a machine learning binary classification framework to differentiate between the diagnosis of ASD and DD. Subsets of 1846 participants were used to train a Random Forest algorithm. Our best classifier achieved, on average, a diagnosis-predicting accuracy of 85.18% (sd 1.11%) on the test samples of 790 participants using 117 genes. The achieved accuracy surpassed results using genetic data and closely matched imaging approaches addressing binary diagnostic classification. Importantly, we observed a similar prediction accuracy when the classifier uses only 62 GO features. This result further corroborates the complex network analysis approach, suggesting that different genetic causes might converge to the dysregulation of the same set of biological mechanisms, leading to a similar disease phenotype. This new biology-driven ontological framework yields a less variable and more compact domain-related set of features with potential mechanistic generalization. The proposed network analysis, allowing for the determination of a clearcut biological distinction between ASD and DD (the latter presenting much lower modularity and heterogeneity), is amenable to machine learning approaches and provides an interesting avenue of research for the future. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

13 pages, 1683 KiB

Open AccessArticle

African Genomic Medicine Portal: A Web Portal for Biomedical Applications

by Houcemeddine Othman, Lyndon Zass, Jorge E. B. da Rocha, Fouzia Radouani, Chaimae Samtal, Ichrak Benamri, Judit Kumuthini, Yasmina J. Fakim, Yosr Hamdi, Nessrine Mezzi, Maroua Boujemaa, Chiamaka Jessica Okeke, Maureen B. Tendwa, Kholoud Sanak, Melek Chaouch, Sumir Panji, Rym Kefi, Reem M. Sallam, Anisah W. Ghoorah, Lilia Romdhane, Anmol Kiran, Ayton P. Meintjes, Perceval Maturure, Haifa Jmel, Ayoub Ksouri, Maryame Azzouzi, Mohammed A. Farahat, Samah Ahmed, Rania Sibira, Michael E. E. Turkson, Alfred Ssekagiri, Ziyaad Parker, Faisal M. Fadlelmola, Kais Ghedira, Nicola Mulder and Samar Kamal Kassim Show full author list Hide full author list

J. Pers. Med. 2022, 12(2), 265; https://doi.org/10.3390/jpm12020265 - 11 Feb 2022

Cited by 1 | Viewed by 4251

Abstract

Genomics data are currently being produced at unprecedented rates, resulting in increased knowledge discovery and submission to public data repositories. Despite these advances, genomic information on African-ancestry populations remains significantly low compared with European- and Asian-ancestry populations. This information is typically segmented across [...] Read more.

Genomics data are currently being produced at unprecedented rates, resulting in increased knowledge discovery and submission to public data repositories. Despite these advances, genomic information on African-ancestry populations remains significantly low compared with European- and Asian-ancestry populations. This information is typically segmented across several different biomedical data repositories, which often lack sufficient fine-grained structure and annotation to account for the diversity of African populations, leading to many challenges related to the retrieval, representation and findability of such information. To overcome these challenges, we developed the African Genomic Medicine Portal (AGMP), a database that contains metadata on genomic medicine studies conducted on African-ancestry populations. The metadata is curated from two public databases related to genomic medicine, PharmGKB and DisGeNET. The metadata retrieved from these source databases were limited to genomic variants that were associated with disease aetiology or treatment in the context of African-ancestry populations. Over 2000 variants relevant to populations of African ancestry were retrieved. Subsequently, domain experts curated and annotated additional information associated with the studies that reported the variants, including geographical origin, ethnolinguistic group, level of association significance and other relevant study information, such as study design and sample size, where available. The AGMP functions as a dedicated resource through which to access African-specific information on genomics as applied to health research, through querying variants, genes, diseases and drugs. The portal and its corresponding technical documentation, implementation code and content are publicly available. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

17 pages, 1083 KiB

Open AccessArticle

SWAAT Bioinformatics Workflow for Protein Structure-Based Annotation of ADME Gene Variants

by Houcemeddine Othman, Sherlyn Jemimah and Jorge Emanuel Batista da Rocha

J. Pers. Med. 2022, 12(2), 263; https://doi.org/10.3390/jpm12020263 - 11 Feb 2022

Cited by 2 | Viewed by 2914

Abstract

Recent genomic studies have revealed the critical impact of genetic diversity within small population groups in determining the way individuals respond to drugs. One of the biggest challenges is to accurately predict the effect of single nucleotide variants and to get the relevant [...] Read more.

Recent genomic studies have revealed the critical impact of genetic diversity within small population groups in determining the way individuals respond to drugs. One of the biggest challenges is to accurately predict the effect of single nucleotide variants and to get the relevant information that allows for a better functional interpretation of genetic data. Different conformational scenarios upon the changing in amino acid sequences of pharmacologically important proteins might impact their stability and plasticity, which in turn might alter the interaction with the drug. Current sequence-based annotation methods have limited power to access this type of information. Motivated by these calls, we have developed the Structural Workflow for Annotating ADME Targets (SWAAT) that allows for the prediction of the variant effect based on structural properties. SWAAT annotates a panel of 36 ADME genes including 22 out of the 23 clinically important members identified by the PharmVar consortium. The workflow consists of a set of Python codes of which the execution is managed within Nextflow to annotate coding variants based on 37 criteria. SWAAT also includes an auxiliary workflow allowing a versatile use for genes other than ADME members. Our tool also includes a machine learning random forest binary classifier that showed an accuracy of 73%. Moreover, SWAAT outperformed six commonly used sequence-based variant prediction tools (PROVEAN, SIFT, PolyPhen-2, CADD, MetaSVM, and FATHMM) in terms of sensitivity and has comparable specificity. SWAAT is available as an open-source tool. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

9 pages, 639 KiB

Open AccessArticle

ICBrainDB: An Integrated Database for Finding Associations between Genetic Factors and EEG Markers of Depressive Disorders

by Roman Ivanov, Fedor Kazantsev, Evgeny Zavarzin, Alexandra Klimenko, Natalya Milakhina, Yury G. Matushkin, Alexander Savostyanov and Sergey Lashin

J. Pers. Med. 2022, 12(1), 53; https://doi.org/10.3390/jpm12010053 - 5 Jan 2022

Cited by 8 | Viewed by 2709

Abstract

In this study, we collected and systemized diverse information related to depressive and anxiety disorders as the first step on the way to investigate the associations between molecular genetics, electrophysiological, behavioral, and psychological characteristics of people. Keeping that in mind, we developed an [...] Read more.

In this study, we collected and systemized diverse information related to depressive and anxiety disorders as the first step on the way to investigate the associations between molecular genetics, electrophysiological, behavioral, and psychological characteristics of people. Keeping that in mind, we developed an internet resource including a database and tools for primary presentation of the collected data of genetic factors, the results of electroencephalography (EEG) tests, and psychological questionnaires. The sample of our study was 1010 people from different regions of Russia. We created the integrated ICBrainDB database that enables users to easily access, download, and further process information about individual behavioral characteristics and psychophysiological responses along with inherited trait data. The data obtained can be useful in training neural networks and in machine learning construction processes in Big Data analysis. We believe that the existence of such a resource will play an important role in the further search for associations of genetic factors and EEG markers of depression. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

15 pages, 324 KiB

Open AccessArticle

Risk-Profile and Feature Selection Comparison in Diabetic Retinopathy

by Valeria Maeda-Gutiérrez, Carlos E. Galván-Tejada, Miguel Cruz, Jorge I. Galván-Tejada, Hamurabi Gamboa-Rosales, Alejandra García-Hernández, Huizilopoztli Luna-García, Irma Gonzalez-Curiel and Mónica Martínez-Acuña

J. Pers. Med. 2021, 11(12), 1327; https://doi.org/10.3390/jpm11121327 - 8 Dec 2021

Cited by 4 | Viewed by 3170

Abstract

One of the main microvascular complications presented in the Mexican population is diabetic retinopathy which affects 27.50% of individuals with type 2 diabetes. Therefore, the purpose of this study is to construct a predictive model to find out the risk factors of this [...] Read more.

One of the main microvascular complications presented in the Mexican population is diabetic retinopathy which affects 27.50% of individuals with type 2 diabetes. Therefore, the purpose of this study is to construct a predictive model to find out the risk factors of this complication. The dataset contained a total of 298 subjects, including clinical and paraclinical features. An analysis was constructed using machine learning techniques including Boruta as a feature selection method, and random forest as classification algorithm. The model was evaluated through a statistical test based on sensitivity, specificity, area under the curve (AUC), and receiving operating characteristic (ROC) curve. The results present significant values obtained by the model obtaining 69% of AUC. Moreover, a risk evaluation was incorporated to evaluate the impact of the predictors. The proposed method identifies creatinine, lipid treatment, glomerular filtration rate, waist hip ratio, total cholesterol, and high density lipoprotein as risk factors in Mexican subjects. The odds ratio increases by 3.5916 times for control patients which have high levels of cholesterol. It is possible to conclude that this proposed methodology is a preliminary computer-aided diagnosis tool for clinical decision-helping to identify the diagnosis of DR. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

7 pages, 997 KiB

Open AccessArticle

Mapping Compound Databases to Disease Maps—A MINERVA Plugin for CandActBase

by Liza Vinhoven, Malte Voskamp and Manuel Manfred Nietert

J. Pers. Med. 2021, 11(11), 1072; https://doi.org/10.3390/jpm11111072 - 24 Oct 2021

Viewed by 2208

Abstract

The MINERVA platform is currently the most widely used platform for visualizing and providing access to disease maps. Disease maps are systems biological maps of molecular interactions relevant in a certain disease context, where they can be used to support drug discovery. For [...] Read more.

The MINERVA platform is currently the most widely used platform for visualizing and providing access to disease maps. Disease maps are systems biological maps of molecular interactions relevant in a certain disease context, where they can be used to support drug discovery. For this purpose, we extended MINERVA’s own drug and chemical search using the MINERVA plugin starter kit. We developed a plugin to provide a linkage between disease maps in MINERVA and application-specific databases of candidate therapeutics. The plugin has three main functionalities; one shows all the targets of all the compounds in the database, the second is a compound-based search to highlight targets of specific compounds, and the third can be used to find compounds that affect a certain target. As a use case, we applied the plugin to link a disease map and compound database we previously established in the context of cystic fibrosis and, herein, point out possible issues and difficulties. The plugin is publicly available on GitLab; the use-case application to cystic fibrosis, connecting disease maps and the compound database CandActCFTR, is available online. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

25 pages, 6729 KiB

Open AccessArticle

A Transcriptomic Analysis of Head and Neck Squamous Cell Carcinomas for Prognostic Indications

by Li-Hsing Chi, Alexander T. H. Wu, Michael Hsiao and Yu-Chuan (Jack) Li

J. Pers. Med. 2021, 11(8), 782; https://doi.org/10.3390/jpm11080782 - 11 Aug 2021

Cited by 6 | Viewed by 4147

Abstract

Survival analysis of the Cancer Genome Atlas (TCGA) dataset is a well-known method for discovering gene expression-based prognostic biomarkers of head and neck squamous cell carcinoma (HNSCC). A cutoff point is usually used in survival analysis for patient dichotomization when using continuous gene [...] Read more.

Survival analysis of the Cancer Genome Atlas (TCGA) dataset is a well-known method for discovering gene expression-based prognostic biomarkers of head and neck squamous cell carcinoma (HNSCC). A cutoff point is usually used in survival analysis for patient dichotomization when using continuous gene expression values. There is some optimization software for cutoff determination. However, the software’s predetermined cutoffs are usually set at the medians or quantiles of gene expression values. There are also few clinicopathological features available in pre-processed datasets. We applied an in-house workflow, including data retrieving and pre-processing, feature selection, sliding-window cutoff selection, Kaplan–Meier survival analysis, and Cox proportional hazard modeling for biomarker discovery. In our approach for the TCGA HNSCC cohort, we scanned human protein-coding genes to find optimal cutoff values. After adjustments with confounders, clinical tumor stage and surgical margin involvement were found to be independent risk factors for prognosis. According to the results tables that show hazard ratios with Bonferroni-adjusted p values under the optimal cutoff, three biomarker candidates, CAMK2N1, CALML5, and FCGBP, are significantly associated with overall survival. We validated this discovery by using the another independent HNSCC dataset (GSE65858). Thus, we suggest that transcriptomic analysis could help with biomarker discovery. Moreover, the robustness of the biomarkers we identified should be ensured through several additional tests with independent datasets. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Graphical abstract

15 pages, 2339 KiB

Open AccessFeature PaperArticle

Identification of Therapeutic Targets for the Selective Killing of HBV-Positive Hepatocytes

by Chien-Jung Huang, Lily Hui-Ching Wang and Yu-Chao Wang

J. Pers. Med. 2021, 11(7), 649; https://doi.org/10.3390/jpm11070649 - 10 Jul 2021

Cited by 1 | Viewed by 2633

Abstract

The hepatitis B virus (HBV) infection is a major risk factor for cirrhosis and hepatocellular carcinoma. Most infected individuals become lifelong carriers of HBV as the drugs currently used to treat the patients can only control the disease, thereby achieving functional cure (loss [...] Read more.

The hepatitis B virus (HBV) infection is a major risk factor for cirrhosis and hepatocellular carcinoma. Most infected individuals become lifelong carriers of HBV as the drugs currently used to treat the patients can only control the disease, thereby achieving functional cure (loss of the hepatitis B surface antigen) but not complete cure (elimination of infected hepatocytes). Therefore, we aimed to identify the target genes for the selective killing of HBV-positive hepatocytes to develop a novel therapy for the treatment of HBV infection. Our strategy was to recognize the conditionally essential genes that are essential for the survival of HBV-positive hepatocytes, but non-essential for the HBV-negative hepatocytes. Using microarray gene expression data curated from the Gene Expression Omnibus database and the known essential genes from the Online GEne Essentiality database, we used two approaches, comprising the random walk with restart algorithm and the support vector machine approach, to determine the potential targets for the selective killing of HBV-positive hepatocytes. The final candidate genes list obtained using these two approaches consisted of 36 target genes, which may be conditionally essential for the cell survival of HBV-positive hepatocytes; however, this requires further experimental validation. Therefore, the genes identified in this study can be used as potential drug targets to develop novel therapeutic strategies for the treatment of HBV, and may ultimately help in achieving the elusive goal of a complete cure for hepatitis B. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

19 pages, 4550 KiB

Open AccessArticle

Exploring the Metabolic Heterogeneity of Cancers: A Benchmark Study of Context-Specific Models

by Mahdi Jalili, Martin Scharm, Olaf Wolkenhauer, Mehdi Damaghi and Ali Salehzadeh-Yazdi

J. Pers. Med. 2021, 11(6), 496; https://doi.org/10.3390/jpm11060496 - 1 Jun 2021

Cited by 12 | Viewed by 4155

Abstract

Metabolic heterogeneity is a hallmark of cancer and can distinguish a normal phenotype from a cancer phenotype. In the systems biology domain, context-specific models facilitate extracting physiologically relevant information from high-quality data. Here, to utilize the heterogeneity of metabolic patterns to discover biomarkers [...] Read more.

Metabolic heterogeneity is a hallmark of cancer and can distinguish a normal phenotype from a cancer phenotype. In the systems biology domain, context-specific models facilitate extracting physiologically relevant information from high-quality data. Here, to utilize the heterogeneity of metabolic patterns to discover biomarkers of all cancers, we benchmarked thousands of context-specific models using well-established algorithms for the integration of omics data into the generic human metabolic model Recon3D. By analyzing the active reactions capable of carrying flux and their magnitude through flux balance analysis, we proved that the metabolic pattern of each cancer is unique and could act as a cancer metabolic fingerprint. Subsequently, we searched for proper feature selection methods to cluster the flux states characterizing each cancer. We employed PCA-based dimensionality reduction and a random forest learning algorithm to reveal reactions containing the most relevant information in order to effectively identify the most influential fluxes. Conclusively, we discovered different pathways that are probably the main sources for metabolic heterogeneity in cancers. We designed the GEMbench website to interactively present the data, methods, and analysis results. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

10 pages, 2363 KiB

Open AccessFeature PaperArticle

Differences among COVID-19, Bronchopneumonia and Atypical Pneumonia in Chest High Resolution Computed Tomography Assessed by Artificial Intelligence Technology

by Robert Chrzan, Monika Bociąga-Jasik, Amira Bryll, Anna Grochowska and Tadeusz Popiela

J. Pers. Med. 2021, 11(5), 391; https://doi.org/10.3390/jpm11050391 - 10 May 2021

Cited by 10 | Viewed by 3614

Abstract

The aim of this study was to compare the results of automatic assessment of high resolution computed tomography (HRCT) by artificial intelligence (AI) in 150 patients from three subgroups: pneumonia in the course of COVID-19, bronchopneumonia and atypical pneumonia. The volume percentage of [...] Read more.

The aim of this study was to compare the results of automatic assessment of high resolution computed tomography (HRCT) by artificial intelligence (AI) in 150 patients from three subgroups: pneumonia in the course of COVID-19, bronchopneumonia and atypical pneumonia. The volume percentage of inflammation and the volume percentage of “ground glass” were significantly higher in the atypical (respectively, 11.04%, 8.61%) and the COVID-19 (12.41%, 10.41%) subgroups compared to the bronchopneumonia (5.12%, 3.42%) subgroup. The volume percentage of consolidation was significantly higher in the COVID-19 (2.95%) subgroup compared to the atypical (1.26%) subgroup. The percentage of “ground glass” in the volume of inflammation was significantly higher in the atypical (89.85%) subgroup compared to the COVID-19 (79.06%) subgroup, which in turn was significantly higher compared to the bronchopneumonia (68.26%) subgroup. HRCT chest images, analyzed automatically by artificial intelligence software, taking into account the structure including “ground glass” and consolidation, significantly differ in three subgroups: COVID-19 pneumonia, bronchopneumonia and atypical pneumonia. However, the partial overlap, particularly between COVID-19 pneumonia and atypical pneumonia, may limit the usefulness of automatic analysis in differentiating the etiology. In our future research, we plan to use artificial intelligence for objective assessment of the dynamics of pulmonary lesions during COVID-19 pneumonia. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

19 pages, 2557 KiB

Open AccessArticle

Predictive Values of Preoperative Characteristics for 30-Day Mortality in Traumatic Hip Fracture Patients

by Yang Cao, Maximilian Peter Forssten, Ahmad Mohammad Ismail, Tomas Borg, Ioannis Ioannidis, Scott Montgomery and Shahin Mohseni

J. Pers. Med. 2021, 11(5), 353; https://doi.org/10.3390/jpm11050353 - 28 Apr 2021

Cited by 17 | Viewed by 2851

Abstract

Hip fracture patients have a high risk of mortality after surgery, with 30-day postoperative rates as high as 10%. This study aimed to explore the predictive ability of preoperative characteristics in traumatic hip fracture patients as they relate to 30-day postoperative mortality using [...] Read more.

Hip fracture patients have a high risk of mortality after surgery, with 30-day postoperative rates as high as 10%. This study aimed to explore the predictive ability of preoperative characteristics in traumatic hip fracture patients as they relate to 30-day postoperative mortality using readily available variables in clinical practice. All adult patients who underwent primary emergency hip fracture surgery in Sweden between 2008 and 2017 were included in the analysis. Associations between the possible predictors and 30-day mortality was performed using a multivariate logistic regression (LR) model; the bidirectional stepwise method was used for variable selection. An LR model and convolutional neural network (CNN) were then fitted for prediction. The relative importance of individual predictors was evaluated using the permutation importance and Gini importance. A total of 134,915 traumatic hip fracture patients were included in the study. The CNN and LR models displayed an acceptable predictive ability for predicting 30-day postoperative mortality using a test dataset, displaying an area under the ROC curve (AUC) of as high as 0.76. The variables with the highest importance in prediction were age, sex, hypertension, dementia, American Society of Anesthesiologists (ASA) classification, and the Revised Cardiac Risk Index (RCRI). Both the CNN and LR models achieved an acceptable performance in identifying patients at risk of mortality 30 days after hip fracture surgery. The most important variables for prediction, based on the variables used in the current study are age, hypertension, dementia, sex, ASA classification, and RCRI. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

15 pages, 4093 KiB

Open AccessArticle

Integrating Patient-Specific Information into Logic Models of Complex Diseases: Application to Acute Myeloid Leukemia

by Alessandro Palma, Marta Iannuccelli, Ilaria Rozzo, Luana Licata, Livia Perfetto, Giorgia Massacci, Luisa Castagnoli, Gianni Cesareni and Francesca Sacco

J. Pers. Med. 2021, 11(2), 117; https://doi.org/10.3390/jpm11020117 - 10 Feb 2021

Cited by 4 | Viewed by 3157

Abstract

High throughput technologies such as deep sequencing and proteomics are increasingly becoming mainstream in clinical practice and support diagnosis and patient stratification. Developing computational models that recapitulate cell physiology and its perturbations in disease is a required step to help with the interpretation [...] Read more.

High throughput technologies such as deep sequencing and proteomics are increasingly becoming mainstream in clinical practice and support diagnosis and patient stratification. Developing computational models that recapitulate cell physiology and its perturbations in disease is a required step to help with the interpretation of results of high content experiments and to devise personalized treatments. As complete cell-models are difficult to achieve, given limited experimental information and insurmountable computational problems, approximate approaches should be considered. We present here a general approach to modeling complex diseases by embedding patient-specific genomics data into actionable logic models that take into account prior knowledge. We apply the strategy to acute myeloid leukemia (AML) and assemble a network of logical relationships linking most of the genes that are found frequently mutated in AML patients. We derive Boolean models from this network and we show that by priming the model with genomic data we can infer relevant patient-specific clinical features. Here we propose that the integration of literature-derived causal networks with patient-specific data should be explored to help bedside decisions. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Graphical abstract

Review

Jump to: Research

15 pages, 1814 KiB

Open AccessReview

Contribution of Synthetic Data Generation towards an Improved Patient Stratification in Palliative Care

by Waldemar Hahn, Katharina Schütte, Kristian Schultz, Olaf Wolkenhauer, Martin Sedlmayr, Ulrich Schuler, Martin Eichler, Saptarshi Bej and Markus Wolfien

J. Pers. Med. 2022, 12(8), 1278; https://doi.org/10.3390/jpm12081278 - 4 Aug 2022

Cited by 6 | Viewed by 3451

Abstract

AI model development for synthetic data generation to improve Machine Learning (ML) methodologies is an integral part of research in Computer Science and is currently being transferred to related medical fields, such as Systems Medicine and Medical Informatics. In general, the idea of [...] Read more.

AI model development for synthetic data generation to improve Machine Learning (ML) methodologies is an integral part of research in Computer Science and is currently being transferred to related medical fields, such as Systems Medicine and Medical Informatics. In general, the idea of personalized decision-making support based on patient data has driven the motivation of researchers in the medical domain for more than a decade, but the overall sparsity and scarcity of data are still major limitations. This is in contrast to currently applied technology that allows us to generate and analyze patient data in diverse forms, such as tabular data on health records, medical images, genomics data, or even audio and video. One solution arising to overcome these data limitations in relation to medical records is the synthetic generation of tabular data based on real world data. Consequently, ML-assisted decision-support can be interpreted more conveniently, using more relevant patient data at hand. At a methodological level, several state-of-the-art ML algorithms generate and derive decisions from such data. However, there remain key issues that hinder a broad practical implementation in real-life clinical settings. In this review, we will give for the first time insights towards current perspectives and potential impacts of using synthetic data generation in palliative care screening because it is a challenging prime example of highly individualized, sparsely available patient information. Taken together, the reader will obtain initial starting points and suitable solutions relevant for generating and using synthetic data for ML-based screenings in palliative care and beyond. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Graphical abstract

24 pages, 2124 KiB

Open AccessEditor’s ChoiceReview

Computational Models for Clinical Applications in Personalized Medicine—Guidelines and Recommendations for Data Integration and Model Validation

by Catherine Bjerre Collin, Tom Gebhardt, Martin Golebiewski, Tugce Karaderi, Maximilian Hillemanns, Faiz Muhammad Khan, Ali Salehzadeh-Yazdi, Marc Kirschner, Sylvia Krobitsch, EU-STANDS4PM consortium and Lars Kuepfer

J. Pers. Med. 2022, 12(2), 166; https://doi.org/10.3390/jpm12020166 - 26 Jan 2022

Cited by 43 | Viewed by 11191

Abstract

The future development of personalized medicine depends on a vast exchange of data from different sources, as well as harmonized integrative analysis of large-scale clinical health and sample data. Computational-modelling approaches play a key role in the analysis of the underlying molecular processes [...] Read more.

The future development of personalized medicine depends on a vast exchange of data from different sources, as well as harmonized integrative analysis of large-scale clinical health and sample data. Computational-modelling approaches play a key role in the analysis of the underlying molecular processes and pathways that characterize human biology, but they also lead to a more profound understanding of the mechanisms and factors that drive diseases; hence, they allow personalized treatment strategies that are guided by central clinical questions. However, despite the growing popularity of computational-modelling approaches in different stakeholder communities, there are still many hurdles to overcome for their clinical routine implementation in the future. Especially the integration of heterogeneous data from multiple sources and types are challenging tasks that require clear guidelines that also have to comply with high ethical and legal standards. Here, we discuss the most relevant computational models for personalized medicine in detail that can be considered as best-practice guidelines for application in clinical care. We define specific challenges and provide applicable guidelines and recommendations for study design, data acquisition, and operation as well as for model validation and clinical translation and other research areas. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

19 pages, 4815 KiB

Open AccessReview

Multi-Layer Picture of Neurodegenerative Diseases: Lessons from the Use of Big Data through Artificial Intelligence

by Andrea Termine, Carlo Fabrizio, Claudia Strafella, Valerio Caputo, Laura Petrosini, Carlo Caltagirone, Emiliano Giardina and Raffaella Cascella

J. Pers. Med. 2021, 11(4), 280; https://doi.org/10.3390/jpm11040280 - 7 Apr 2021

Cited by 28 | Viewed by 5350

Abstract

In the big data era, artificial intelligence techniques have been applied to tackle traditional issues in the study of neurodegenerative diseases. Despite the progress made in understanding the complex (epi)genetics signatures underlying neurodegenerative disorders, performing early diagnosis and developing drug repurposing strategies remain [...] Read more.

In the big data era, artificial intelligence techniques have been applied to tackle traditional issues in the study of neurodegenerative diseases. Despite the progress made in understanding the complex (epi)genetics signatures underlying neurodegenerative disorders, performing early diagnosis and developing drug repurposing strategies remain serious challenges for such conditions. In this context, the integration of multi-omics, neuroimaging, and electronic health records data can be exploited using deep learning methods to provide the most accurate representation of patients possible. Deep learning allows researchers to find multi-modal biomarkers to develop more effective and personalized treatments, early diagnosis tools, as well as useful information for drug discovering and repurposing in neurodegenerative pathologies. In this review, we will describe how relevant studies have been able to demonstrate the potential of deep learning to enhance the knowledge of neurodegenerative disorders such as Alzheimer’s and Parkinson’s diseases through the integration of all sources of biomedical data. Full article

(This article belongs to the Special Issue Systems Medicine and Bioinformatics)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Systems Medicine and Bioinformatics

Share This Special Issue

Special Issue Editors

Special Issue Information

Keywords

Benefits of Publishing in a Special Issue

Published Papers (15 papers)

Research

Review

Further Information

Guidelines

MDPI Initiatives

Follow MDPI