Vincenzo Lagani

Frailty phenotypes in the elderly based on cluster analysis: a longitudinal study of two Danish cohorts. Evidence for a genetic influence on frailty

Frailty is a physiological state characterized by the deregulation of multiple physiologic systems of an aging organism determining the loss of homeostatic capacity, which exposes the elderly to disability, diseases, and finally death. An operative definition of frailty, useful for the classification of the individual quality of aging, is needed. On the other hand, the documented heterogeneity in the quality of aging among different geographic areas suggests the necessity for a frailty classification approach providing population-specific results. Moreover, the contribution of the individual genetic background on the frailty status is still questioned. We investigated the applicability of a cluster analysis approach based on specific geriatric parameters, previously set up and validated in a southern Italian population, to two large longitudinal Danish samples. In both cohorts, we identified groups of subjects homogeneous for their frailty status and characterized by different survival patterns. A subsequent survival analysis availing of Accelerated Failure Time models allowed us to formulate an operative index able to correlate classification variables with survival probability. From these models, we quantified the differential effect of various parameters on survival, and we estimated the heritability of the frailty phenotype by exploiting the twin pairs in our sample. These data suggest the presence of a genetic influence on the frailty variability and indicate that cluster analysis can define specific frailty phenotypes in each population.

A systematic review of predictive risk models for diabetes complications based on large scale clinical studies.

This work presents a systematic review of long-term risk assessment models for evaluating the probability of developing complications in diabetes patients. Diabetes mellitus can cause many complications if not adequately controlled; risk assessment models can help physicians and patients in identifying the complications most likely to arise and in taking the necessary countermeasures. We identified six large medical studies related to diabetes mellitus upon which current available risk assessment models are built on; all these studies had duration over 5 years and most of them included some common demographic and clinical data strongly related to diabetic complications. The most common predictions for long term diabetes complications are related to cardiovascular diseases and diabetic retinopathy. Our analysis of the literature led us to the conclusion that researchers and medical practitioners should take in account that some limitations undermine the applicability of risk assessment models; for example, it is hard to judge whether results obtained on a specific cohort can be effectively translated to other populations. Nevertheless, all these studies have significantly contributed to identify significant risk factors associated with the major diabetes complications.

Structure-based variable selection for survival data

MOTIVATION Variable selection is a typical approach used for molecular-signature and biomarker discovery; however, its application to survival data is often complicated by censored samples. We propose a new algorithm for variable selection suitable for the analysis of high-dimensional, right-censored data called Survival Max-Min Parents and Children (SMMPC). The algorithm is conceptually simple, scalable, based on the theory of Bayesian networks (BNs) and the Markov blanket and extends the corresponding algorithm (MMPC) for classification tasks. The selected variables have a structural interpretation: if T is the survival time (in general the time-to-event), SMMPC returns the variables adjacent to T in the BN representing the data distribution. The selected variables also have a causal interpretation that we discuss. RESULTS We conduct an extensive empirical analysis of prototypical and state-of-the-art variable selection algorithms for survival data that are applicable to high-dimensional biological data. SMMPC selects on average the smallest variable subsets (less than a dozen per dataset), while statistically significantly outperforming all of the methods in the study returning a manageable number of genes that could be inspected by a human expert. AVAILABILITY Matlab and R code are freely available from http://www.mensxmachina.org

Meta-analysis of clinical metabolic profiling studies in cancer: challenges and opportunities.

Cancer cell metabolism has received increasing attention. Despite a boost in the application of clinical metabolic profiling (CMP) in cancer patients, a meta‐analysis has not been performed. The primary goal of this study was to assess whether public accessibility of metabolomics data and identification and reporting of metabolites were sufficient to assess which metabolites were consistently altered in cancer patients. We therefore retrospectively curated data from CMP studies in cancer patients published during 5 recent years and used an established vote‐counting method to perform a semiquantitative meta‐analysis of metabolites in tumor tissue and blood. This analysis confirmed well‐known increases in glycolytic metabolites, but also unveiled unprecedented changes in other metabolites such as ketone bodies and amino acids (histidine, tryptophan). However, this study also highlighted that insufficient public accessibility of metabolomics data, and inadequate metabolite identification and reporting hamper the discovery potential of meta‐analyses of CMP studies, calling for improved standardization of metabolomics studies.

Performance-Estimation Properties of Cross-Validation-Based Protocols with Simultaneous Hyper-Parameter Optimization

In a typical supervised data analysis task, one needs to perform the following two tasks: (a) select the best combination of learning methods (e.g., for variable selection and classifier) and tune their hyper-parameters (e.g., K in K-NN), also called model selection, and (b) provide an estimate of the performance of the final, reported model. Combining the two tasks is not trivial because when one selects the set of hyper-parameters that seem to provide the best estimated performance, this estimation is optimistic (biased / overfitted) due to performing multiple statistical comparisons. In this paper, we confirm that the simple Cross-Validation with model selection is indeed optimistic (overestimates) in small sample scenarios. In comparison the Nested Cross Validation and the method by Tibshirani and Tibshirani provide conservative estimations, with the later protocol being more computationally efficient. The role of stratification of samples is examined and it is shown that stratification is beneficial.

Contribution of genetic polymorphisms on functional status at very old age: A gene-based analysis of 38 genes (311 SNPs) in the oxidative stress pathway

Preservation of functional ability is a well-recognized marker of longevity. At a molecular level, a major determinant of the physiological decline occurring with aging is the imbalance between production and accumulation of oxidative damage to macromolecules, together with a decreased efficiency of stress response to avoid or repair such damage. In this paper we investigated the association of 38 genes (311 SNPs) belonging to the pro-antioxidant pathways with physical and cognitive performances, by analyzing single SNP and gene-based associations with Hand Grip strength (HG), Activities of Daily Living (ADL), Walking Speed (WS), Mini Mental State Examination (MMSE) and Composite Cognitive Score (CCS) in a Cohort of 1089 Danish nonagenarians. Moreover, for each gene analyzed in the pro-antioxidant pathway, we tested the influence on longitudinal survival. In the whole sample, nominal associations were found for TXNRD1 variability with ADL and WS, NDUFS1 and UCP3 with HG and WS, GCLC and UCP2 with WS (p<0.05). Stronger associations although not holding the multiple comparison correction, were observed between MMSE and NDUFV1, MT1A and GSTP1 variability (p<0.009). Moreover, we found that association between genetic variability in the pro-antioxidant pathway and functional status at old age is influenced by sex. In particular, most significant associations were observed in nonagenarian females, between HG scores and GLRX and UCP3 variability, between ADL levels and TXNRD1, MMSE and MT1A genetic variability. In males, a borderline statistically significant association with ADL level was found for UQCRFS1 gene. Nominally significant associations in relation to survival were found in the female sample only with SOD2, NDUFS1, UCP3 and TXNRD1 variability, the latter two confirming previous observations reported in the same cohort. Overall, our work supports the evidence that genes belonging to the pro-anti-oxidant pathway are able to modulate physical and cognitive performance after the ninth decade of life, finally influencing extreme survival.

Development and validation of risk assessment models for diabetes-related complications based on the DCCT/EDIC data

AIM To derive and validate a set of computational models able to assess the risk of developing complications and experiencing adverse events for patients with diabetes. The models are developed on data from the Diabetes Control and Complications Trial (DCCT) and the Epidemiology of Diabetes Interventions and Complications (EDIC) studies, and are validated on an external, retrospectively collected cohort. METHODS We selected fifty-one clinical parameters measured at baseline during the DCCT as potential risk factors for the following adverse outcomes: Cardiovascular Diseases (CVD), Hypoglycemia, Ketoacidosis, Microalbuminuria, Proteinuria, Neuropathy and Retinopathy. For each outcome we applied a data-mining analysis protocol in order to identify the best-performing signature, i.e., the smallest set of clinical parameters that, considered jointly, are maximally predictive for the selected outcome. The predictive models built on the selected signatures underwent both an interval validation on the DCCT/EDIC data and an external validation on a retrospective cohort of 393 diabetes patients (49 Type I and 344 Type II) from the Chorleywood Medical Center, UK. RESULTS The selected predictive signatures contain five to fifteen risk factors, depending on the specific outcome. Internal validation performances, as measured by the Concordance Index (CI), range from 0.62 to 0.83, indicating good predictive power. The models achieved comparable performances for the Type I and, quite surprisingly, Type II external cohort. CONCLUSIONS Data-mining analyses of the DCCT/EDIC data allow the identification of accurate predictive models for diabetes-related complications. We also present initial evidences that these models can be applied on a more recent, European population.

UCP3 polymorphisms, hand grip performance and survival at old age: Association analysis in two Danish middle aged and elderly cohorts

An efficient uncoupling process is generally considered to have a protective effect on the aging muscle by slowing down its age-related decay. Genetic polymorphisms in the Uncoupling Protein 3 (UCP3) gene, whose product is mainly expressed in skeletal muscle, were suggested to be associated with hand grip (HG) performances in elderly populations. Considering the population specificity of the quality of aging, we aimed to add further support to this evidence by analyzing the association between four SNPs in the UCP3 gene and relative haplotypes in two large cohorts of middle aged (N=708) and oldest old Danes (N=908). We found that the variability at rs1685354 and rs11235972 was associated with HG levels both at single and haplotypic level in both cohorts. Furthermore, taking advantage of large cohort and period survival data of the oldest cohort, we tested the association of each SNP with survival at 10years from the baseline visit. Interestingly, we found that allele A at rs11235972, associated in this cohort with lowest HG scores, influences also the survival patterns, with people carrying this allele showing higher mortality rates. On the whole, our work supports the role of UCP3 gene in functional status and survival at old age.

Probabilistic Computational Causal Discovery for Systems Biology

Discovering the causal mechanisms of biological systems is necessary to design new drugs and therapies. Computational Causal Discovery (CD) is a field that offers the potential to discover causal relations and causal models under certain conditions with a limited set of interventions/manipulations. This chapter reviews the basic concepts and principles of CD, the nature of the assumptions to enable it, potential pitfalls in its application, and recent advances and directions. Importantly, several success stories in molecular and systems biology are discussed in detail.

Hidden treasures in "ancient" microarrays: gene-expression portrays biology and potential resistance pathways of major lung cancer subtypes and normal tissue.

Objective: Novel statistical methods and increasingly more accurate gene annotations can transform “old” biological data into a renewed source of knowledge with potential clinical relevance. Here, we provide an in silico proof-of-concept by extracting novel information from a high-quality mRNA expression dataset, originally published in 2001, using state-of-the-art bioinformatics approaches. Methods: The dataset consists of histologically defined cases of lung adenocarcinoma (AD), squamous (SQ) cell carcinoma, small-cell lung cancer, carcinoid, metastasis (breast and colon AD), and normal lung specimens (203 samples in total). A battery of statistical tests was used for identifying differential gene expressions, diagnostic and prognostic genes, enriched gene ontologies, and signaling pathways. Results: Our results showed that gene expressions faithfully recapitulate immunohistochemical subtype markers, as chromogranin A in carcinoids, cytokeratin 5, p63 in SQ, and TTF1 in non-squamous types. Moreover, biological information with putative clinical relevance was revealed as potentially novel diagnostic genes for each subtype with specificity 93–100% (AUC = 0.93–1.00). Cancer subtypes were characterized by (a) differential expression of treatment target genes as TYMS, HER2, and HER3 and (b) overrepresentation of treatment-related pathways like cell cycle, DNA repair, and ERBB pathways. The vascular smooth muscle contraction, leukocyte trans-endothelial migration, and actin cytoskeleton pathways were overexpressed in normal tissue. Conclusion: Reanalysis of this public dataset displayed the known biological features of lung cancer subtypes and revealed novel pathways of potentially clinical importance. The findings also support our hypothesis that even old omics data of high quality can be a source of significant biological information when appropriate bioinformatics methods are used.