APDB: a database on air pollutant characterization and similarity prediction.

The World Health Organization estimates that 9 out of 10 people worldwide breathe air containing high levels of pollutants. Long-term and chronic exposure to high concentrations of air pollutants is associated with deleterious effects on vital organs, including increased inflammation in the lungs, oxidative stress in the heart and disruption of the blood-brain barrier. For this reason, in an effort to find an association between exposure to pollutants and the toxicological effects observable on human health, an online resource collecting and characterizing in detail pollutant molecules could be helpful to investigate their properties and mechanisms of action. We developed a database, APDB, collecting air-pollutant-related data from different online resources, in particular, molecules from the US Environmental Protection Agency, their associated targets and bioassays found in the PubChem chemical repository and their computed molecular descriptors and quantum mechanics properties. A web interface allows (i) to browse data by category, (ii) to navigate the database by querying molecules and targets and (iii) to visualize and download molecule and target structures as well as computed descriptors and similarities. The desired data can be freely exported in textual/tabular format and the whole database in SQL format. Database URL http://apdb.di.univr.it.

KUALA: a machine learning-driven framework for kinase inhibitors repositioning.

The family of protein kinases comprises more than 500 genes involved in numerous functions. Hence, their physiological dysfunction has paved the way toward drug discovery for cancer, cardiovascular, and inflammatory diseases. As a matter of fact, Kinase binding sites high similarity has a double role. On the one hand it is a critical issue for selectivity, on the other hand, according to poly-pharmacology, a synergistic controlled effect on more than one target could be of great pharmacological interest. Another important aspect of binding similarity is the possibility of exploit it for repositioning of drugs on targets of the same family. In this study, we propose our approach called Kinase drUgs mAchine Learning frAmework (KUALA) to automatically identify kinase active ligands by using specific sets of molecular descriptors and provide a multi-target priority score and a repurposing threshold to suggest the best repurposable and non-repurposable molecules. The comprehensive list of all kinase-ligand pairs and their scores can be found at https://github.com/molinfrimed/multi-kinases .

PI3K-driven HER2 expression is a potential therapeutic target in colorectal cancer stem cells.

OBJECTIVE: Cancer stem cells are responsible for tumour spreading and relapse. Human epidermal growth factor receptor 2 (HER2) expression is a negative prognostic factor in colorectal cancer (CRC) and a potential target in tumours carrying the gene amplification. Our aim was to define the expression of HER2 in colorectal cancer stem cells (CR-CSCs) and its possible role as therapeutic target in CRC resistant to anti- epidermal growth factor receptor (EGFR) therapy. DESIGN: A collection of primary sphere cell cultures obtained from 60 CRC specimens was used to generate CR-CSC mouse avatars to preclinically validate therapeutic options. We also made use of the ChIP-seq analysis for transcriptional evaluation of HER2 activation and global RNA-seq to identify the mechanisms underlying therapy resistance. RESULTS: Here we show that in CD44v6-positive CR-CSCs, high HER2 expression levels are associated with an activation of the phosphatidylinositol 3-kinase (PI3K)/AKT pathway, which promotes the acetylation at the regulatory elements of the Erbb2 gene. HER2 targeting in combination with phosphatidylinositol 3-kinase (PI3K) and mitogen-activated protein kinase kinase (MEK) inhibitors induces CR-CSC death and regression of tumour xenografts, including those carrying Kras and Pik3ca mutation. Requirement for the triple targeting is due to the presence of cancer-associated fibroblasts, which release cytokines able to confer CR-CSC resistance to PI3K/AKT inhibitors. In contrast, targeting of PI3K/AKT as monotherapy is sufficient to kill liver-disseminating CR-CSCs in a model of adjuvant therapy. CONCLUSIONS: While PI3K targeting kills liver-colonising CR-CSCs, the concomitant inhibition of PI3K, HER2 and MEK is required to induce regression of tumours resistant to anti-EGFR therapies. These data may provide a rationale for designing clinical trials in the adjuvant and metastatic setting.

Choline PET/CT features to predict survival outcome in high-risk prostate cancer restaging: a preliminary machine-learning radiomics study.

BACKGROUND: Radiomic features are increasingly utilized to evaluate tumor heterogeneity in PET imaging but to date its role has not been investigated for Cho-PET in prostate cancer. The potential application of radiomics features analysis using a machine-learning radiomics algorithm was evaluated to select 18F-Cho PET/CT imaging features to predict disease progression in PCa. METHODS: We retrospectively analyzed high-risk PCa patients who underwent restaging 18F-Cho PET/CT from November 2013 to May 2018. 18F-Cho PET/CT studies and related structures containing volumetric segmentations were imported in the "CGITA" toolbox to extract imaging features from each lesion. A Machine-learning model has been adapted using NCA for feature selection, while DA was used as a method for feature classification and performance analysis. RESULTS: One hundred and six imaging features were extracted for 46 lesions for a total of 4876 features analyzed. No significant differences between the training and validating sets in terms of age, sex, PSA values, lesion location and size (P>0.05) were demonstrated by the machine-learning model. Thirteen features were able to discriminate FU disease status after NCA selection. Best performance in DA classification was obtained using the combination of the 13 selected features (sensitivity 74%, specificity 58% and accuracy 66%) compared to the use of all features (sensitivity 40%, specificity 52%, and accuracy 51%). Per-site performance of the 13 selected features in DA classification were as follows: T = sensitivity 63%, specificity 83%, accuracy 71%; N = sensitivity 87%, specificity 91% of and accuracy 90%; bone-M = sensitivity 33%, specificity 77% and accuracy 66%. CONCLUSIONS: An artificial intelligence model demonstrated to be feasible and able to select a panel of 18F-Cho PET/CT features with valuable association with PCa patients' outcome.

Exploring Machine Learning Techniques to Predict the Response to Omalizumab in Chronic Spontaneous Urticaria.

BACKGROUND: Omalizumab is the best treatment for patients with chronic spontaneous urticaria (CSU). Machine learning (ML) approaches can be used to predict response to therapy and the effectiveness of a treatment. No studies are available on the use of ML techniques to predict the response to Omalizumab in CSU. METHODS: Data from 132 CSU outpatients were analyzed. Urticaria Activity Score over 7 days (UAS7) and treatment efficacy were assessed. Clinical and demographic characteristics were used for training and validating ML models to predict the response to treatment. Two methodologies were used to label the data based on the response to treatment (UAS7 ≥ 6): (A) at 1, 3 and 5 months; (B) classifying the patients as early responders (ER), late responders (LR) or non-responders (NR) (ER: UAS 7 ≥ 6 at first month, LR: UAS 7 ≥ 6 at third month, NR: if none of the previous conditions occurred). RESULTS: ER were predominantly characterized by hypertension, while LR mainly suffered from asthma and hypothyroidism. A slight positive correlation (R2 = 0.21) was found between total IgE levels and UAS7 at 1 month. Variable Importance Analysis (VIA) reported D-dimer and C-reactive proteins as the key blood tests for the performance of learning techniques. Using methodology (A), SVM (specificity of 0.81) and k-NN (sensitivity of 0.8) are the best models to predict LR at the third month. CONCLUSION: k-NN plus the SVM model could be used to identify the response to treatment. D-dimer and C-reactive proteins have greater predictive power in training ML models.

Adipose stem cell niche reprograms the colorectal cancer stem cell metastatic machinery.

Obesity is a strong risk factor for cancer progression, posing obesity-related cancer as one of the leading causes of death. Nevertheless, the molecular mechanisms that endow cancer cells with metastatic properties in patients affected by obesity remain unexplored.Here, we show that IL-6 and HGF, secreted by tumor neighboring visceral adipose stromal cells (V-ASCs), expand the metastatic colorectal (CR) cancer cell compartment (CD44v6 + ), which in turn secretes neurotrophins such as NGF and NT-3, and recruits adipose stem cells within tumor mass. Visceral adipose-derived factors promote vasculogenesis and the onset of metastatic dissemination by activation of STAT3, which inhibits miR-200a and enhances ZEB2 expression, effectively reprogramming CRC cells into a highly metastatic phenotype. Notably, obesity-associated tumor microenvironment provokes a transition in the transcriptomic expression profile of cells derived from the epithelial consensus molecular subtype (CMS2) CRC patients towards a mesenchymal subtype (CMS4). STAT3 pathway inhibition reduces ZEB2 expression and abrogates the metastatic growth sustained by adipose-released proteins. Together, our data suggest that targeting adipose factors in colorectal cancer patients with obesity may represent a therapeutic strategy for preventing metastatic disease.

PNPLA3 and TLL-1 Polymorphisms as Potential Predictors of Disease Severity in Patients With COVID-19.

Albeit the pathogenesis of COVID-19 remains unclear, host's genetic polymorphisms in genes involved in infection and reinfection, inflammation, or immune stimulation could play a role in determining the course and outcome. We studied in the early phase of pandemic consecutive patients (N = 383) with SARS-CoV-2 infection, whose subsequent clinical course was classified as mild or severe, the latter being characterized by admission to intensive therapy unit or death. Five host gene polymorphisms (MERTK rs4374383, PNPLA3 rs738409, TLL-1 rs17047200, IFNL3 rs1297860, and INFL4 rs368234815) were assessed by using whole nucleic acids extracted from nasopharyngeal swabs. Specific protease cleavage sites of TLL-1 on the SARS-CoV-2 Spike protein were predicted in silico. Male subjects and older patients were significantly at higher risk for a severe outcome (p = 0.02 and p < 0.001, respectively). By considering patients ≤65 years, after adjusting for potential confounding due to sex, an increased risk of severe outcome was found in subjects with the GG genotype of PNPLA3 (adj-OR: 4.69; 95% CI = 1.01-22.04) or TT genotype of TLL-1 (adj-OR=9.1; 95% CI = 1.45-57.3). In silico evaluation showed that TLL-1 is potentially involved in the Spike protein cleavage which is essential for viral binding and entry into the host cells using the host receptor angiotensin-converting enzyme 2 (ACE2). Subjects carrying a GG genotype in PNPLA3 gene might have a constitutive upregulation of the NLRP3 inflammasome and be more prone to tissue damage when infected by SARS-CoV-2. The TT genotype in TLL-1 gene might affect its protease activity on the SARS-CoV-2 Spike protein, enhancing the ability to infect or re-infect host's cells. The untoward effect of these variants on disease course is evident in younger patients due to the relative absence of comorbidities as determinants of prognosis. In the unresolved pathogenetic scenery of COVID-19, the identification of genetic variants associates with more prolonged course or with a severe outcome of infection would support the development of predictive tools useful to stratify subjects by risk class at presentation. Moreover, the individuation of key genes could contribute to a better understanding of the pathways involved in the pathogenesis, giving the basis for rational therapeutic approaches.

Targeting chemoresistant colorectal cancer via systemic administration of a BMP7 variant.

Despite intense research and clinical efforts, patients affected by advanced colorectal cancer (CRC) have still a poor prognosis. The discovery of colorectal (CR) cancer stem cell (CSC) as the cell compartment responsible for tumor initiation and propagation may provide new opportunities for the development of new therapeutic strategies. Given the reduced sensitivity of CR-CSCs to chemotherapy and the ability of bone morphogenetic proteins (BMP) to promote colonic stem cell differentiation, we aimed to investigate whether an enhanced variant of BMP7 (BMP7v) could sensitize to chemotherapy-resistant CRC cells and tumors. Thirty-five primary human cultures enriched in CR-CSCs, including four from chemoresistant metastatic lesions, were used for in vitro studies and to generate CR-CSC-based mouse avatars to evaluate tumor growth and progression upon treatment with BMP7v alone or in combination with standard therapy or PI3K inhibitors. BMP7v treatment promotes CR-CSC differentiation and recapitulates the cell differentiation-related gene expression profile by suppressing Wnt pathway activity and reducing mesenchymal traits and survival of CR-CSCs. Moreover, in CR-CSC-based mouse avatars, BMP7v exerts an antiangiogenic effect and sensitizes tumor cells to standard chemotherapy regardless of the mutational, MSI, and CMS profiles. Of note, tumor harboring PIK3CA mutations were affected to a lower extent by the combination of BMP7v and chemotherapy. However, the addition of a PI3K inhibitor to the BMP7v-based combination potentiates PIK3CA-mutant tumor drug response and reduces the metastatic lesion size. These data suggest that BMP7v treatment may represent a useful antiangiogenic and prodifferentiation agent, which renders CSCs sensitive to both standard and targeted therapies.

18F-Florbetaben PET/CT to Assess Alzheimer's Disease: A new Analysis Method for Regional Amyloid Quantification.

BACKGROUND AND PURPOSE: While AD can be definitively confirmed by postmortem histopathologic examination, in vivo imaging may improve the clinician's ability to identify AD at the earliest stage. The aim of the study was to test the performance of amyloid PET using new processing imaging algorithm for more precise diagnosis of AD. METHODS: Amyloid PET results using a new processing imaging algorithm (MRI-Less and AAL Atlas) were correlated with clinical, cognitive status, CSF analysis, and other imaging. The regional SUVR using the white matter of cerebellum as reference region and scores from clinical and cognitive tests were used to create ROC curves. Leave-one-out cross-validation was carried out to validate the results. RESULTS: Forty-four consecutive patients with clinical evidence of dementia, were retrospectively evaluated. Amyloid PET scan was positive in 26/44 patients with dementia. After integration with 18F-FDG PET, clinical data and CSF protein levels, 22 of them were classified as AD, the remaining 4 as vascular or frontotemporal dementia. Amyloid and FDG PET, CDR 1, CSF Tau, and p-tau levels showed the best true positive and true negative rates (amyloid PET: AUC = .85, sensitivity .91, specificity .79). A SUVR value of 1.006 in the inferior frontal cortex and of 1.03 in the precuneus region was the best cutoff SUVR value and showed a good correlation with the diagnosis of AD. Thirteen of 44 amyloid PET positive patients have been enrolled in clinical trials using antiamyloid approaches. CONCLUSIONS: Amyloid PET using SPM-normalized SUVR analysis showed high predictive power for the differential diagnosis of AD.

INBIA: a boosting methodology for proteomic network inference.

BACKGROUND: The analysis of tissue-specific protein interaction networks and their functional enrichment in pathological and normal tissues provides insights on the etiology of diseases. The Pan-cancer proteomic project, in The Cancer Genome Atlas, collects protein expressions in human cancers and it is a reference resource for the functional study of cancers. However, established protocols to infer interaction networks from protein expressions are still missing. RESULTS: We have developed a methodology called Inference Network Based on iRefIndex Analysis (INBIA) to accurately correlate proteomic inferred relations to protein-protein interaction (PPI) networks. INBIA makes use of 14 network inference methods on protein expressions related to 16 cancer types. It uses as reference model the iRefIndex human PPI network. Predictions are validated through non-interacting and tissue specific PPI networks resources. The first, Negatome, takes into account likely non-interacting proteins by combining both structure properties and literature mining. The latter, TissueNet and GIANT, report experimentally verified PPIs in more than 50 human tissues. The reliability of the proposed methodology is assessed by comparing INBIA with PERA, a tool which infers protein interaction networks from Pathway Commons, by both functional and topological analysis. CONCLUSION: Results show that INBIA is a valuable approach to predict proteomic interactions in pathological conditions starting from the current knowledge of human protein interactions.

A novel computational method for inferring competing endogenous interactions.

Posttranscriptional cross talk and communication between genes mediated by microRNA response element (MREs) yield large regulatory competing endogenous RNA (ceRNA) networks. Their inference may improve the understanding of pathologies and shed new light on biological mechanisms. A variety of RNA: messenger RNA, transcribed pseudogenes, noncoding RNA, circular RNA and proteins related to RNA-induced silencing complex complex interacting with RNA transfer and ribosomal RNA have been experimentally proved to be ceRNAs. We retrace the ceRNA hypothesis of posttranscriptional regulation from its original formulation [Salmena L, Poliseno L, Tay Y, et al. Cell 2011;146:353-8] to the most recent experimental and computational validations. We experimentally analyze the methods in literature [Li J-H, Liu S, Zhou H, et al. Nucleic Acids Res 2013;42:D92-7; Sumazin P, Yang X, Chiu H-S, et al. Cell 2011;147:370-81; Sarver AL, Subramanian S. Bioinformation 2012;8:731-3] comparing them with a general machine learning approach, called ceRNA predIction Algorithm, evaluating the performance in predicting novel MRE-based ceRNAs.