Evaluation of Some Statistical Methods for the Identification of Differentially Expressed Genes

Microarray platforms have been around for many years and while there is a rise of new technologies in laboratories, microarrays are still prevalent. When it comes to the analysis of microarray data to identify differentially expressed (DE) genes, many methods have been proposed and modified for improvement. However, the most popular methods such as Significance Analysis of Microarrays (SAM), samroc, fold change, and rank product are far from perfect. When it comes down to choosing which method is most powerful, it comes down to the characteristics of the sample and distribution of the gene expressions. The most practiced method is usually SAM or samroc but when the data tends to be skewed, the power of these methods decrease. With the concept that the median becomes a better measure of central tendency than the mean when the data is skewed, the tests statistics of the SAM and fold change methods are modified in this thesis. This study shows that the median modified fold change method improves the power for many cases when identifying DE genes if the data follows a lognormal distribution.

Biochemical and Structural Studies on PrfA, the Transcriptional Regulator of Virulence in Listeria monocytogenes

Abstract

Listeria monocytogenes is a gram-positive soil saprophytic bacterium that is capable of causing fatal infection in humans. The main virulence regulator PrfA, a member of the Crp/FNR family of transcriptional regulators, activates the expression of essential proteins required for host cell invasion and cell-to-cell spread. The mechanism of PrfA activation and the identity of its small molecule coactivator have remained a mystery for more than 20 years, but it is hypothesized that PrfA shares mechanistic similarity to the E. coli cAMP binding protein, Crp. Crp activates gene expression by binding cAMP, increasing the DNA binding affinity of the protein and causing a significant DNA bend that facilitates RNA polymerase binding and downstream gene activation. Our data suggests PrfA activates virulence protein expression through a mechanism distinct from the canonical Crp activation mechanism that involves a combination of cysteine residue reduction and glutathione (GSH) binding.

Listeria lacking glutathione synthase (ΔgshF) is avirulent in mice; however virulence is rescued when the bacterium expresses the constitutively active PrfA mutant G145S. Interestingly, Listeria expressing a PrfA mutant in which its four cysteines are mutated to alanine (Quad PrfA), demonstrate a 30-fold decrease in virulence. The Quad and ΔgshF double mutant strains are avirulent. DNA-binding affinity, measured through fluorescence polarization assays, indicate reduction of the cysteine side chains is sufficient to allow PrfA to binds its physiological promoters Phly and PactA with low nanomolar affinity. Oxidized PrfA binds the promoters poorly.

Unexpectedly, Quad also binds promoter DNA with nanomolar affinity, suggesting that the cysteines play a role in transcription efficiency in addition to DNA binding. Both PrfA and Quad bind GSH at physiologically relevant and comparable affinities, however GSH did not affect DNA binding in either case. Thermal denaturation assays suggest that Quad and wild-type PrfA differ structurally upon binding GSH, which supports the in vivo difference in infection between the regulator and its mutant.

Structures of PrfA in complex with cognate DNA, determined through X-ray crystallography, further support the disparity between PrfA and Crp activation mechanisms as two structures of reduced PrfA bound to Phly (PrfA-Phly30 and PrfA-Phly24) suggest the DNA adopts a less bent DNA conformation when compared to Crp-cAMP- DNA. The structure of Quad-Phly30 confirms the DNA-binding data as the protein-DNA complex adopts the same overall conformation as PrfA-Phly.

From these results, we hypothesize a two-step activation mechanism wherein PrfA, oxidized upon cell entry and unable to bind DNA, is reduced upon its intracellular release and binds DNA, causing a slight bend in the promoter and small increase in transcription of PrfA-regulated genes. The structures of PrfA-Phly30 and PrfA-Phly24 likely visualize this intermediate complex. Increasing concentrations of GSH shift the protein to a (PrfA-GSH)-DNA complex which is fully active transcriptionally and is hypothesized to resemble closely the transcriptionally active structure of the cAMP-(Crp)-DNA complex. Thermal denaturation results suggest Quad PrfA is deficient in this second step, which explains the decrease in virulence and implicates the cysteine residues as critical for transcription efficiency. Further structural and biochemical studies are on-going to clarify this mechanism of activation.

Unraveling the function of bacterial-type phosphoenolpyruvate carboxylase in vascular plants

Two distinct phosphoenolpyruvate carboxylase (PEPC) isozymes occur in vascular plants and green algae: plant-type PEPC (PTPC) and bacterial-type PEPC (BTPC). PTPC polypeptides typically form a tightly regulated cytosolic Class-1 PEPC homotetramer. BTPCs, however, appear to be less widely expressed and to exist only as catalytic and regulatory subunits that physically interact with co-expressed PTPC subunits to form hetero-octameric Class-2 PEPC complexes that are highly desensitized to Class-1 PEPC allosteric effectors. Yeast two-hybrid studies indicated that castor plant BTPC (RcPPC4) interacts with all three Arabidopsis thaliana PTPC isozymes, and that it forms stronger interactions with AtPPC2 and AtPPC3, suggesting that specific PTPCs are preferred for Class-2 PEPC formation. In contrast, Arabidopsis BTPC (AtPPC4) appeared to interact very weakly with AtPPC2 and AtPPC3, suggesting that BTPCs from different species may have different physical properties, hypothesized to be due to sequence dissimilarities within their ~10 kDa intrinsically disordered region. Recent RNA-seq and microarray data were analyzed to obtain a better understanding of BTPC expression patterns in different tissues of various monocot and dicot species. High levels of BTPC transcripts, polypeptides and Class-2 PEPC complexes were originally discovered in developing castor seeds, but the analysis revealed a broad range of diverse tissues where abundant BTPC transcripts are also expressed, such as the developing fruits of cucumber, grape, and tomato. Marked BTPC expression correlated well with the presence of ~116 kDa immunoreactive BTPC polypeptides, as well as Class-2 PEPC complexes in the immature fruit of cucumbers and tomatoes. It is therefore hypothesized that in vascular plants BTPC and thus Class-2 PEPC complexes maintain anaplerotic PEP flux in tissues with elevated malate levels that would potently inhibit ‘housekeeping’ Class-1 PEPCs. Elevated levels of malate can be used by biosynthetically active sink tissues such as immature tomatoes and cucumbers for rapid cell expansion, drought or salt stressed roots for osmoregulation, and developing seeds and pollen as a precursor for storage lipid and protein biosynthesis.

Characterisation of a secretory serine protease inhibitor (SjB6) from Schistosoma japonicum

BACKGROUND: Proteins belonging to the serine protease inhibitor (serpin) superfamily play essential physiological roles in many organisms. In pathogens, serpins are thought to have evolved specifically to limit host immune responses by interfering with the host immune-stimulatory signals. Serpins are less well characterised in parasitic helminths, although some are thought to be involved in mechanisms associated with host immune modulation. In this study, we cloned and partially characterised a secretory serpin from Schistosoma japonicum termed SjB6, these findings provide the basis for possible functional roles.

METHODS: SjB6 gene was identified through database mining of our previously published microarray data, cloned and detailed sequence and structural analysis and comparative modelling carried out using various bioinformatics and proteomics tools. Gene transcriptional profiling was determined by real-time PCR and the expression of native protein determined by immunoblotting. An immunological profile of the recombinant protein produced in insect cells was determined by ELISA.

RESULTS: SjB6 contains an open reading frame of 1160 base pairs that encodes a protein of 387 amino acid residues. Detailed sequence analysis, comparative modelling and structural-based alignment revealed that SjB6 contains the essential structural motifs and consensus secondary structures typical of inhibitory serpins. The presence of an N-terminal signal sequence indicated that SjB6 is a secretory protein. Real-time data indicated that SjB6 is expressed exclusively in the intra-mammalian stage of the parasite life cycle with its highest expression levels in the egg stage (p < 0.0001). The native protein is approximately 60 kDa in size and recombinant SjB6 (rSjB6) was recognised strongly by sera from rats experimentally infected with S. japonicum.

CONCLUSIONS: The significantly high expression of SjB6 in schistosome eggs, when compared to other life cycle stages, suggests a possible association with disease pathology, while the strong reactivity of sera from experimentally infected rats against rSjB6 suggests that native SjB6 is released into host tissue and induces an immune response. This study presents a comprehensive demonstration of sequence and structural-based analysis of a secretory serpin from a trematode and suggests SjB6 may be associated with important functional roles in S. japonicum, particularly in parasite modulation of the host microenvironment.

Targeting treatment resistant breast cancer stem cells with FKBPL and its peptide derivative, AD-01

FKBPL and its peptide derivative, AD-01, have already demonstrated well-established inhibitory effects on breast cancer growth and CD44 dependent anti-angiogenic activity1, 2, 3. Since breast cancer stem cells (BCSCs) are CD44 positive, we wanted to explore if AD-01 could specifically target BCSCs. FKBPL stable overexpression or AD-01 treatment were highly effective at reducing the BCSC population measured by inhibiting mammosphere forming efficiency (MFE) in cell lines and primary breast cancer samples from both solid breast tumours and pleural effusions. Flow cytometry, to assess the ESA+/CD44+/CD24- subpopulation, validated these results. The ability of AD-01 to inhibit the self-renewal capacity of BCSCs was confirmed across three generations of mammospheres, where mammospheres were completely eradicated by the third generation (p<0.001). Clonogenic assays suggested that AD-01 mediated BCSC differentiation, with a significant decrease in the number of holoclones and an associated increase in meroclones/paraclones. In support of this, the stem cell markers, Nanog and Oct4 were significantly reduced following AD-01 treatment, whilst transfection of FKBPL-targeted siRNAs led to an increase in these markers and in mammosphere forming potential, highlighting the endogenous role of FKBPL in stem cell signalling. The clinical relevance of this was confirmed using a publically available microarray data set (GSE7390), where, high FKBPL and low Nanog expression were independently associated with improved overall survival in breast cancer patients (log rank test p=0.03; hazard ratio=3.01). When AD-01 was combined with other agents, we observed synergistic activity with the Notch inhibitor, DAPT and AD-01 was also able to abrogate a chemo- and radiotherapy induced enrichment in BCSCs. Importantly, using ‘gold standard’ in vivo limiting dilution assays we demonstrated a delay in tumour initiation and reoccurrence in AD-01 treated xenografts. In summary, AD-01 appears to have dual anti-angiogenic and anti-BCSC activity which will be advantageous as this agent enters clinical trial.

Targeting treatment resistant cancer stem cells with FKBPL and its peptide therapeutics

FKBPL and its peptide derivatives have already demonstrated well-established inhibitory effects on cancer growth and CD44-dependent anti-angiogenic activity. Since cancer stem cells (CSCs) are CD44 positive, we wanted to explore if these therapeutics could specifically target CSCs in breast and ovarian cancer. In a tumoursphere assay, FKBPL stable overexpression or FKBPL-based peptide (AD-01, preclinical peptide or ALM201, clinical peptide candidate) treatment were highly effective at reducing the CSC population measured by inhibiting tumoursphere forming efficiency in breast and ovarian cancer cell lines and primary breast cancer samples from both solid breast tumours and pleural effusions. Flow cytometry, to assess the ESA+/CD44+/CD24- and ALDH+ cell subpopulations representative of CSCs, validated these results. The ability of AD-01 and ALM201 to inhibit the self-renewal capacity of CSCs was confirmed across three generations, eradicating CSC completely by the third generation (p<0.001). Furthermore, clonogenic assay demonstrated that FKBPL-based peptides mediated CSC differentiation, with a significant decrease in the number of CSCs or holoclones and an associated increase in differentiated cancer cells or meroclones/paraclones. In addition, AD-01 treatment in vitro and in vivo led to a significant reduction in the stem cell markers, Nanog, Sox2 and Oct4 protein and mRNA levels; whilst transfection of FKBPL-targeted siRNAs led to an increase in these markers and in tumoursphere forming potential, highlighting the endogenous role of FKBPL in stem cell signalling. The clinical relevance of this was confirmed using a publically available microarray data set (GSE7390), where, high FKBPL and low Nanog expression were independently associated with improved overall survival in breast cancer patients (log rank test p=0.03; hazard ratio=3.01). Additionally, when AD-01 was combined with other agents, we observed additive activity with the Notch inhibitor, DAPT and AD-01 was also able to abrogate a chemo- and radiotherapy induced enrichment in CSCs. Importantly, using gold standard in vivo limiting dilution assays we demonstrated a delay in tumour initiation and reoccurrence in AD-01 treated xenografts. In summary, FKBPL-based peptides appear to have dual anti-angiogenic and anti-CSC activity which will be advantageous as this agent enters clinical trial.

Caractérisation des effets de l’hyperglycémie chronique dans la choroïde d’yeux diabétiques

But: Le diabète est un problème important de santé publique et la complication oculaire la plus commune est la rétinopathie diabétique (RD). Certaines études indiquent que la choroïde des patients diabétiques est affectée sans signe apparent de RD. Notre hypothèse est que l’élévation du stress oxydant liée à l’hyperglycémie chronique affecte la fonction choroïdienne à un stade précoce de la RD. Nous proposons d’étudier la glycolyse, le métabolisme mitochondrial, le stress nitrosatif et la méthylation de l’ADN ainsi que de caractériser les modifications histologiques dans la choroïde diabétique. Méthodes: L’expression des gènes/protéines associés à la glycolyse, au métabolisme mitochondrial et à la production de l’oxyde nitrique a été comparée par profilage génique et immunobuvardage Western entre les choroïdes saines et diabétiques. Les niveaux globaux de méthylation et d’hydroxyméthylation de l’ADN ont été quantifiés par immunoslot blot et HPLC-MS/MS dans ces tissus. Enfin, des coupes tissulaires d’yeux de donneurs sains ou diabétiques avec RD non proliférante (RDNP) ou proliférante (RDP) ont été colorées au trichrome de Masson et au Weigert. L’épaisseur de la choroïde et de la membrane de Bruch, ainsi que la densité et le diamètre des vaisseaux sanguins choroïdiens ont été analysés. Résultats: Nos résultats montrent une dérégulation de l’expression de certains transcrits de la choroïde diabétique, mais peu de différences au niveau de l’expression protéique des cibles validées. Le niveau global de méthylation de l’ADN est similaire entre les donneurs sains et diabétiques. Nos analyses histologiques démontrent une diminution de l’épaisseur de la choroïde et une dégénérescence des choriocapillaires et des veines/veinules chez les donneurs diabétiques atteints de RDP. Conclusions: L’étude de la choroïde est importante, car l’atteinte de ce tissu a de graves répercussions sur la fonction rétinienne. L’identification de cibles dans la choroïde ouvre de nouvelles perspectives pour un traitement préventif de la RD.

TRIB2 in human AML: a biological and clinical investigation

Acute myeloid leukemia (AML) involves the proliferation, abnormal survival and arrest of cells at a very early stage of myeloid cell differentiation. The biological and clinical heterogeneity of this disease complicates treatment and highlights the significance of understanding the underlying causes of AML, which may constitute potential therapeutic targets, as well as offer prognostic information. Tribbles homolog 2 (Trib2) is a potent murine oncogene capable of inducing transplantable AML with complete penetrance. The pathogenicity of Trib2 is attributed to its ability to induce proteasomal degradation of the full length isoform of the transcription factor CCAAT/enhancer-binding protein alpha (C/EBPα p42). The role of TRIB2 in human AML cells, however, has not been systematically investigated or targeted. Across human cancers, TRIB2 oncogenic activity was found to be associated with its elevated expression. In the context of AML, TRIB2 overexpression was suggested to be associated with the large and heterogeneous subset of cytogenetically normal AML patients. Based upon the observation that overexpression of TRIB2 has a role in cellular transformation, the effect of modulating its expression in human AML was examined in a human AML cell line that expresses high levels of TRIB2, U937 cells. Specific suppression of TRIB2 led to impaired cell growth, as a consequence of both an increase in apoptosis and a decrease in cell proliferation. Consistent with these in vitro results, TRIB2 silencing strongly reduced progression of the U937 in vivo xenografts, accompanied by detection of a lower spleen weight when compared with mice transplanted with TRIB2- expressing control cells. Gene expression analysis suggested that TRIB2 modulates apoptosis and cell-cycle sensitivity by influencing the expression of a subset of genes known to have implications on these phenotypes. Furthermore, TRIB2 was found to be expressed in a significant subset of AML patient samples analysed. To investigate whether increased expression of this gene could be afforded prognostic significance, primary AML cells with dichotomized levels of TRIB2 transcripts were evaluated in terms of their xenoengraftment potential, an assay reported to correlate with disease aggressiveness observed in humans. A small cohort of analysed samples with higher TRIB2 expression did not associate with preferential leukaemic cell engraftment in highly immune-deficient mice, hence, not predicting for an adverse prognosis. However, further experiments including a larger cohort of well characterized AML patients would be needed to clarify TRIB2 significance in the diagnostic setting. Collectively, these data support a functional role for TRIB2 in the maintenance of the oncogenic properties of human AML cells and suggest TRIB2 can be considered a rational therapeutic target. Proteasome inhibition has emerged as an attractive target for the development of novel anti-cancer therapies and results from translational research and clinical trials support the idea that proteasome inhibitors should be considered in the treatment of AML. The present study argued that proteasome inhibition would effectively inhibit the function of TRIB2 by abrogating C/EBPα p42 protein degradation and that it would be an effective pharmacological targeting strategy in TRIB2-positive AMLs. Here, a number of cell models expressing high levels of TRIB2 were successfully targeted by treatment with proteasome inhibitors, as demonstrated by multiple measurements that included increased cytotoxicity, inhibition of clonogenic growth and anti-AML activity in vivo. Mechanistically, it was shown that block of the TRIB2 degradative function led to an increase of C/EBPα p42 and that response was specific to the TRIB2-C/EBPα axis. Specificity was addressed by a panel of experiments showing that U937 cells (express detectable levels of endogenous TRIB2 and C/EBPα) treated with the proteasome inhibitor bortezomib (Brtz) displayed a higher cytotoxic response upon TRIB2 overexpression and that ectopic expression of C/EBPα rescued cell death. Additionally, in C/EBPα-negative leukaemia cells, K562 and Kasumi 1, Brtz-induced toxicity was not increased following TRIB2 overexpression supporting the specificity of the compound on the TRIB2-C/EBPα axis. Together these findings provide pre-clinical evidence that TRIB2- expressing AML cells can be pharmacologically targeted with proteasome inhibition due, in part, to blockage of the TRIB2 proteolytic function on C/EBPα p42. A large body of evidence indicates that AML arises through the stepwise acquisition of genetic and epigenetic changes. Mass spectrometry data has identified an interaction between TRIB2 and the epigenetic regulator Protein Arginine Methyltransferase 5 (PRMT5). Following assessment of TRIB2‟s role in AML cell survival and effective targeting of the TRIB2-C/EBPα degradation pathway, a putative TRIB2/PRMT5 cooperation was investigated in order to gain a deeper understanding of the molecular network in which TRIB2 acts as a potent myeloid oncogene. First, a microarray data set was interrogated for PRMT5 expression levels and the primary enzyme responsible for symmetric dimethylation was found to be transcribed at significantly higher levels in AML patients when compared to healthy controls. Next, depletion of PRMT5 in the U937 cell line was shown to reduce the transformative phenotype in the high expressing TRIB2 AML cells, which suggests that PRMT5 and TRIB2 may cooperate to maintain the leukaemogenic potential. Importantly, PRMT5 was identified as a TRIB2-interacting protein by means of a protein tagging approach to purify TRIB2 complexes from 293T cells. These findings trigger further research aimed at understanding the underlying mechanism and the functional significance of this interplay. In summary, the present study provides experimental evidence that TRIB2 has an important oncogenic role in human AML maintenance and, importantly in such a molecularly heterogeneous disease, provides the rational basis to consider proteasome inhibition as an effective targeting strategy for AML patients with high TRIB2 expression. Finally, the identification of PRMT5 as a TRIB2-interacting protein opens a new level of regulation to consider in AML. This work may contribute to our further understanding and therapeutic strategies in acute leukaemias.

Transcriptional profiling in Saccharomyces cerevisiae relevant for predicting alachlor mechanisms of toxicity

Alachlor has been a commonly applied herbicide and is a substance of ecotoxicological concern. The present study aims to identify molecular biomarkers in the eukaryotic model Saccharomyces cerevisiae that can be used to predict potential cytotoxic effects of alachlor, while providing new mechanistic clues with possible relevance for experimentally less accessible eukaryotes. It focuses on genome-wide expression profiling in a yeast population in response to two exposure scenarios exerting effects from slight to moderate magnitude at phenotypic level. In particular, 100 and 264 genes, respectively, were found as differentially expressed on a 2-h exposure of yeast cells to the lowest observed effect concentration (110 mg/L) and the 20% inhibitory concentration (200 mg/L) of alachlor, in comparison with cells not exposed to the herbicide. The datasets of alachlor-responsive genes showed functional enrichment in diverse metabolic, transmembrane transport, cell defense, and detoxification categories. In general, the modifications in transcript levels of selected candidate biomarkers, assessed by quantitative reverse transcriptase polymerase chain reaction, confirmed the microarray data and varied consistently with the growth inhibitory effects of alachlor. Approximately 16% of the proteins encoded by alachlor-differentially expressed genes were found to share significant homology with proteins from ecologically relevant eukaryotic species. The biological relevance of these results is discussed in relation to new insights into the potential adverse effects of alachlor in health of organisms from ecosystems, particularly in worst-case situations such as accidental spills or careless storage, usage, and disposal.

Unraveling the function of bacterial-type phosphoenolpyruvate carboxylase in vascular plants

Thesis (Master, Biology) -- Queen's University, 2016-09-29 20:09:46.997

Policitemia Vera: scoperta di saggi di biopsia liquida finalizzata a identificare i biomarcatori correlati al rischio trombotico

La Policitemia Vera (PV) è una neoplasia mieloproliferativa con un aumentato rischio di trombosi e di progressione verso la Mielofibrosi. L'infiammazione cronica è comunemente osservata nelle neoplasie mieloproliferative, compresa la PV. La rete infiammatoria, tra le varie componenti, comprende le vescicole extracellulari (EVs), che svolgono un ruolo nella comunicazione cellula-cellula. Inoltre, le componenti microbiche circolanti sono state recentemente indicate come potenziali modificatori dell'infiammazione, della coagulazione e dell’emopoiesi in generale. Qui abbiamo studiato il DNA microbico delle EVs circolanti attraverso. Sangue periferico e feci sono stati raccolti da pazienti con PV (n=38) e da donatori sani (n=30). Le EVs circolanti derivate da megacariociti (MK) e piastrine (PLT) sono state analizzate mediante citometria a flusso. Dopo l'estrazione del DNA microbico dalle feci e dalle EV isolate, è stata sequenziata la regione V3-V4 del 16S rDNA. La percentuale di EVs di MK era ridotta nei pazienti con PV rispetto ai donatori sani. Al contrario, la proporzione di EVs di PLT era aumentata. La PV è stata associata anche a una firma del DNA microbico delle EVs isolate con una maggiore diversità e una composizione microbica distinta rispetto alla controparte sana. Nei pazienti con PV c’è una maggiore proporzione di EVs associate al lipopolisaccaride. Il profilo del microbioma intestinale non differiva tra PV e doantori. Inoltre, l'aumento della proporzione di EVs di MK e la riduzione di EVs di piastre identificavano i pazienti con pregressa trombosi. Le EVs dei pazienti con trombosi erano impoverite di DNA di Staphylococcus ma arricchite di DNA di Actinobacteria e Anaerococcus. Inoltre, questi pazienti avevano livelli più bassi di EVs associate al lipopolisaccaride. I pazienti con fibrosi midollare avevano una maggiore proporzione di PE-EV ed erano arricchite in DNA di Collinsella e Flavobacterium. Questi dati possono contribuire a perfezionare la prognosi della PV e a identificare nuovi bersagli farmacologici.

Marcatori di eta biologica nella malattia di Parkinson: risultati di un'analisi glicomica e epigenomica

La malattia di Parkinson è una malattia neurodegenerativa caratterizzata da una progressiva disfunzione motoria e cognitiva. È noto che l'età avanzata è il principale fattore di rischio per la malattia di Parkinson e alcuni studi hanno dimostrato un'accelerazione dell'età biologica nelle fasi più avanzate della malattia. Questo studio si propone di valutare se l'accelerazione dell'invecchiamento biologico descritta nelle fasi avanzate della malattia di Parkinson caratterizzi anche le prime fasi della malattia. A tal fine sono stati utilizzati due tipi di marcatori biologici di età, basati sull'analisi della metilazione del DNA del sangue (l'orologio epigenetico e sue varianti) e dei profili degli N-glicani nel plasma (GlycoAge Test). I biomarcatori sono stati valutati in campioni ottenuti da pazienti con malattia di Parkinson de novo, con diagnosi recente e non ancora in trattamento farmacologico, nonché da pazienti con stadi più avanzati della malattia e da controlli sani. I risultati ottenuti nelle prime fasi della malattia non mostrano segni di invecchiamento accelerato, che trovano conferma nelle fasi più avanzate. Dai dati di metilazione è possibile prevedere le proporzioni delle diverse popolazioni di leucociti. Questa analisi nelle prime fasi della malattia ha già evidenziato significative alterazioni che seguono in parte quelle caratteristiche dell'invecchiamento del sistema immunitario, suggerendo un'immunosenescenza accelerata nella malattia di Parkinson. Infine, i dati sulla metilazione del DNA sono stati analizzati per identificare le differenze nelle regioni metilate del genoma tra pazienti con malattia di Parkinson e controlli. I risultati suggeriscono l'esistenza di piccole ma significative alterazioni nella metilazione del DNA che caratterizzano lo stadio precoce e/o avanzato della malattia. In conclusione, questo studio suggerisce che le prime fasi della malattia di Parkinson sono caratterizzate da specifiche alterazioni epigenetiche e invecchiamento precoce del sistema immunitario, che tuttavia non si traducono in un'alterazione dei biomarcatori di invecchiamento epigenetici e glicomici.

Selection bias in gene extraction on the basis of microarray gene-expression data

In the context of cancer diagnosis and treatment, we consider the problem of constructing an accurate prediction rule on the basis of a relatively small number of tumor tissue samples of known type containing the expression data on very many (possibly thousands) genes. Recently, results have been presented in the literature suggesting that it is possible to construct a prediction rule from only a few genes such that it has a negligible prediction error rate. However, in these results the test error or the leave-one-out cross-validated error is calculated without allowance for the selection bias. There is no allowance because the rule is either tested on tissue samples that were used in the first instance to select the genes being used in the rule or because the cross-validation of the rule is not external to the selection process; that is, gene selection is not performed in training the rule at each stage of the cross-validation process. We describe how in practice the selection bias can be assessed and corrected for by either performing a cross-validation or applying the bootstrap external to the selection process. We recommend using 10-fold rather than leave-one-out cross-validation, and concerning the bootstrap, we suggest using the so-called. 632+ bootstrap error estimate designed to handle overfitted prediction rules. Using two published data sets, we demonstrate that when correction is made for the selection bias, the cross-validated error is no longer zero for a subset of only a few genes.

A mixture model-based approach to the clustering of microarray expression data

Motivation: This paper introduces the software EMMIX-GENE that has been developed for the specific purpose of a model-based approach to the clustering of microarray expression data, in particular, of tissue samples on a very large number of genes. The latter is a nonstandard problem in parametric cluster analysis because the dimension of the feature space (the number of genes) is typically much greater than the number of tissues. A feasible approach is provided by first selecting a subset of the genes relevant for the clustering of the tissue samples by fitting mixtures of t distributions to rank the genes in order of increasing size of the likelihood ratio statistic for the test of one versus two components in the mixture model. The imposition of a threshold on the likelihood ratio statistic used in conjunction with a threshold on the size of a cluster allows the selection of a relevant set of genes. However, even this reduced set of genes will usually be too large for a normal mixture model to be fitted directly to the tissues, and so the use of mixtures of factor analyzers is exploited to reduce effectively the dimension of the feature space of genes. Results: The usefulness of the EMMIX-GENE approach for the clustering of tissue samples is demonstrated on two well-known data sets on colon and leukaemia tissues. For both data sets, relevant subsets of the genes are able to be selected that reveal interesting clusterings of the tissues that are either consistent with the external classification of the tissues or with background and biological knowledge of these sets.