arXiv subjects

Alex Graudenzi

Publications and source records attributed to Alex Graudenzi.

At least 19 recordsLinked to original sources

arXiv2022

Unity is strength: Improving the Detection of Adversarial Examples with Ensemble Approaches

A key challenge in computer vision and deep learning is the definition of robust strategies for the detection of adversarial examples. Here, we propose the adoption of ensemble approaches to leverage the effectiveness of multiple detectors in exploiting distinct properties of the input data. To this end, the ENsemble Adversarial Detector (ENAD) framework integrates scoring functions from state-of-the-art detectors based on Mahalanobis distance, Local Intrinsic Dimensionality, and One-Class Support Vector Machines, which process the hidden features of deep neural networks. ENAD is designed to ensure high standardization and reproducibility to the computational workflow. Importantly, extensive tests on benchmark datasets, models and adversarial attacks show that ENAD outperforms all competing methods in the large majority of settings. The improvement over the state-of-the-art and the intrinsic generality of the framework, which allows one to easily extend ENAD to include any set of detectors, set the foundations for the new area of ensemble adversarial detection.

cs.CV↗

arXiv2021

OG-SPACE: Optimized Stochastic Simulation of Spatial Models of Cancer Evolution

Algorithmic strategies for the spatio-temporal simulation of multi-cellular systems are crucial to generate synthetic datasets for bioinformatics tools benchmarking, as well as to investigate experimental hypotheses on real-world systems in a variety of in-silico scenarios. In particular, efficient algorithms are needed to overcome the harsh trade-off between scalability and expressivity, which typically limits our capability to produce realistic simulations, especially in the context of cancer evolution. We introduce the Optimized Gillespie algorithm for simulating Stochastic sPAtial models of Cancer Evolution (OG-SPACE), a computational framework for the simulation of the spatio-temporal evolution of cancer subpopulations and of the experimental procedures of both bulk andsingle-cell sequencing. OG-SPACE relies on an evolution of the Gillespie algorithm optimized to deal with large numbers of cells and is designed tohandle a variety of birth-death processes and interaction rules on arbitrary lattices. As output OG-SPACE returns: the visual snapshots of the spatial configuration of the system over time, the phylogeny of the (sampled) cells, the mutational tree, the variant allele frequency spectrum (for bulk experiments) and the cell genotypes (for single-cell experiments).OG-SPACE is freely available at:https://github.com/BIMIB-DISCo/OG-SPACE

q-bio.PE↗

arXiv2021

PMCE: efficient inference of expressive models of cancer evolution with high prognostic power

Motivation: Driver (epi)genomic alterations underlie the positive selection of cancer subpopulations, which promotes drug resistance and relapse. Even though substantial heterogeneity is witnessed in most cancer types, mutation accumulation patterns can be regularly found and can be exploited to reconstruct predictive models of cancer evolution. Yet, available methods cannot infer logical formulas connecting events to represent alternative evolutionary routes or convergent evolution. Results: We introduce PMCE, an expressive framework that leverages mutational profiles from cross-sectional sequencing data to infer probabilistic graphical models of cancer evolution including arbitrary logical formulas, and which outperforms the state-of-the-art in terms of accuracy and robustness to noise, on simulations. The application of PMCE to 7866 samples from the TCGA database allows us to identify a highly significant correlation between the predicted evolutionary paths and the overall survival in 7 tumor types, proving that our approach can effectively stratify cancer patients in reliable risk groups. Availability: PMCE is freely available at https://github.com/BIMIB-DISCo/PMCE, in addition to the code to replicate all the analyses presented in the manuscript. Contacts: daniele.ramazzotti@unimib.it, alex.graudenzi@ibfm.cnr.it.

stat.ML↗

arXiv2020

Investigating the Compositional Structure Of Deep Neural Networks

The current understanding of deep neural networks can only partially explain how input structure, network parameters and optimization algorithms jointly contribute to achieve the strong generalization power that is typically observed in many real-world applications. In order to improve the comprehension and interpretability of deep neural networks, we here introduce a novel theoretical framework based on the compositional structure of piecewise linear activation functions. By defining a direct acyclic graph representing the composition of activation patterns through the network layers, it is possible to characterize the instances of the input data with respect to both the predicted label and the specific (linear) transformation used to perform predictions. Preliminary tests on the MNIST dataset show that our method can group input instances with regard to their similarity in the internal representation of the neural network, providing an intuitive measure of input complexity.

cs.LG↗

arXiv2019

cyTRON and cyTRON/JS: two Cytoscape-based applications for the inference of cancer evolution models

The increasing availability of sequencing data of cancer samples is fueling the development of algorithmic strategies to investigate tumor heterogeneity and infer reliable models of cancer evolution. We here build up on previous works on cancer progression inference from genomic alteration data, to deliver two distinct Cytoscape-based applications, which allow to produce, visualize and manipulate cancer evolution models, also by interacting with public genomic and proteomics databases. In particular, we here introduce cyTRON, a stand-alone Cytoscape app, and cyTRON/JS, a web application which employs the functionalities of Cytoscape/JS. cyTRON was developed in Java; the code is available at https://github.com/BIMIB-DISCo/cyTRON and on the Cytoscape App Store http://apps.cytoscape.org/apps/cytron. cyTRON/JS was developed in JavaScript and R; the source code of the tool is available at https://github.com/BIMIB-DISCo/cyTRON-js and the tool is accessible from https://bimib.disco.unimib.it/cytronjs/welcome.

q-bio.GN↗

arXiv2019

Learning mutational graphs of individual tumour evolution from single-cell and multi-region sequencing data

Background. A large number of algorithms is being developed to reconstruct evolutionary models of individual tumours from genome sequencing data. Most methods can analyze multiple samples collected either through bulk multi-region sequencing experiments or the sequencing of individual cancer cells. However, rarely the same method can support both data types. Results. We introduce TRaIT, a computational framework to infer mutational graphs that model the accumulation of multiple types of somatic alterations driving tumour evolution. Compared to other tools, TRaIT supports multi-region and single-cell sequencing data within the same statistical framework, and delivers expressive models that capture many complex evolutionary phenomena. TRaIT improves accuracy, robustness to data-specific errors and computational complexity compared to competing methods. Conclusions. We show that the application of TRaIT to single-cell and multi-region cancer datasets can produce accurate and reliable models of single-tumour evolution, quantify the extent of intra-tumour heterogeneity and generate new testable experimental hypotheses.

q-bio.GN↗

arXiv2018

Efficient computational strategies to learn the structure of probabilistic graphical models of cumulative phenomena

Structural learning of Bayesian Networks (BNs) is a NP-hard problem, which is further complicated by many theoretical issues, such as the I-equivalence among different structures. In this work, we focus on a specific subclass of BNs, named Suppes-Bayes Causal Networks (SBCNs), which include specific structural constraints based on Suppes' probabilistic causation to efficiently model cumulative phenomena. Here we compare the performance, via extensive simulations, of various state-of-the-art search strategies, such as local search techniques and Genetic Algorithms, as well as of distinct regularization methods. The assessment is performed on a large number of simulated datasets from topologies with distinct levels of complexity, various sample size and different rates of errors in the data. Among the main results, we show that the introduction of Suppes' constraints dramatically improve the inference accuracy, by reducing the solution space and providing a temporal ordering on the variables. We also report on trade-offs among different search techniques that can be efficiently employed in distinct experimental settings. This manuscript is an extended version of the paper "Structural Learning of Probabilistic Graphical Models of Cumulative Phenomena" presented at the 2018 International Conference on Computational Science.

cs.LG↗

arXiv2018

Modeling cumulative biological phenomena with Suppes-Bayes Causal Networks

Several diseases related to cell proliferation are characterized by the accumulation of somatic DNA changes, with respect to wildtype conditions. Cancer and HIV are two common examples of such diseases, where the mutational load in the cancerous/viral population increases over time. In these cases, selective pressures are often observed along with competition, cooperation and parasitism among distinct cellular clones. Recently, we presented a mathematical framework to model these phenomena, based on a combination of Bayesian inference and Suppes' theory of probabilistic causation, depicted in graphical structures dubbed Suppes-Bayes Causal Networks (SBCNs). SBCNs are generative probabilistic graphical models that recapitulate the potential ordering of accumulation of such DNA changes during the progression of the disease. Such models can be inferred from data by exploiting likelihood-based model-selection strategies with regularization. In this paper we discuss the theoretical foundations of our approach and we investigate in depth the influence on the model-selection task of: (i) the poset based on Suppes' theory and (ii) different regularization strategies. Furthermore, we provide an example of application of our framework to HIV genetic data highlighting the valuable insights provided by the inferred.

cs.AI↗

arXiv2017

Metabolic enrichment through functional gene rules

It is well known that tumors originating from the same tissue have different prognosis and sensitivity to treatments. Over the last decade, cancer genomics consortia like the Cancer Genome Atlas (TCGA) have been generating thousands of cross-sectional data, for thousands of human primary tumors originated from various tissues. Thanks to that public database, it is today possible to analyze a broad range of relevant information such as gene sequences, expression profiles or metabolite footprints, to capture tumor molecular heterogeneity and improve patient stratification and clinical management. To this aim, it is common practice to analyze datasets grouped into clusters based on clinical observations and/or molecular features. However, the identification of specific properties of each cluster that may be effectively targeted by therapeutic drugs still represents a challenging task. We define a method to generate an activity score for the metabolic reactions of different clusters of patients based on their transcriptional profile. This approach reduces the number of variables from many genes to few reactions, by aggregating transcriptional information associated to the same enzymatic reaction according to gene-enzyme and enzyme-reaction rules. We also applied the methodology to a dataset of 244 RNAseq transcriptional profiles taken from patients with colorectal cancer (CRC). CRC samples are typically divided into two sub-types: (i) tumors with microsatellite instability (MSI), associated with hyper-mutation and with CpG island methylation phenotype, and (ii) microsatellite stable (MSS) tumors, typically endowed with chromosomal instability. We report some key differences in the central carbon metabolism of the two clusters. We also show how the method can be used to describe the metabolism of individual patients and cluster them exclusively based on metabolic features.

q-bio.CB↗

arXiv2017

Algorithmic Methods to Infer the Evolutionary Trajectories in Cancer Progression

The genomic evolution inherent to cancer relates directly to a renewed focus on the voluminous next generation sequencing (NGS) data, and machine learning for the inference of explanatory models of how the (epi)genomic events are choreographed in cancer initiation and development. However, despite the increasing availability of multiple additional -omics data, this quest has been frustrated by various theoretical and technical hurdles, mostly stemming from the dramatic heterogeneity of the disease. In this paper, we build on our recent works on "selective advantage" relation among driver mutations in cancer progression and investigate its applicability to the modeling problem at the population level. Here, we introduce PiCnIc (Pipeline for Cancer Inference), a versatile, modular and customizable pipeline to extract ensemble-level progression models from cross-sectional sequenced cancer genomes. The pipeline has many translational implications as it combines state-of-the-art techniques for sample stratification, driver selection, identification of fitness-equivalent exclusive alterations and progression model inference. We demonstrate PiCnIc's ability to reproduce much of the current knowledge on colorectal cancer progression, as well as to suggest novel experimentally verifiable hypotheses.

q-bio.GN↗

arXiv2016

TRONCO: an R package for the inference of cancer progression models from heterogeneous genomic data

Motivation: We introduce TRONCO (TRanslational ONCOlogy), an open-source R package that implements the state-of-the-art algorithms for the inference of cancer progression models from (epi)genomic mutational profiles. TRONCO can be used to extract population-level models describing the trends of accumulation of alterations in a cohort of cross-sectional samples, e.g., retrieved from publicly available databases, and individual-level models that reveal the clonal evolutionary history in single cancer patients, when multiple samples, e.g., multiple biopsies or single-cell sequencing data, are available. The resulting models can provide key hints in uncovering the evolutionary trajectories of cancer, especially for precision medicine or personalized therapy. Availability: TRONCO is released under the GPL license, it is hosted in the Software section at http://bimib.disco.unimib.it/ and archived also at bioconductor.org. Contact: tronco@disco.unimib.it

q-bio.QM↗

arXiv2015

CABeRNET: a Cytoscape app for Augmented Boolean models of gene Regulatory NETworks

Background. Dynamical models of gene regulatory networks (GRNs) are highly effective in describing complex biological phenomena and processes, such as cell differentiation and cancer development. Yet, the topological and functional characterization of real GRNs is often still partial and an exhaustive picture of their functioning is missing. Motivation. We here introduce CABeRNET, a Cytoscape app for the generation, simulation and analysis of Boolean models of GRNs, specifically focused on their augmentation when a only partial topological and functional characterization of the network is available. By generating large ensembles of networks in which user-defined entities and relations are added to the original core, CABeRNET allows to formulate hypotheses on the missing portions of real networks, as well to investigate their generic properties, in the spirit of complexity science. Results. CABeRNET offers a series of innovative simulation and modeling functions and tools, including (but not being limited to) the dynamical characterization of the gene activation patterns ruling cell types and differentiation fates, and sophisticated robustness assessments, as in the case of gene knockouts. The integration within the widely used Cytoscape framework for the visualization and analysis of biological networks, makes CABeRNET a new essential instrument for both the bioinformatician and the computational biologist, as well as a computational support for the experimentalist. An example application concerning the analysis of an augmented T-helper cell GRN is provided.

q-bio.MN↗

arXiv2015

CAPRI: Efficient Inference of Cancer Progression Models from Cross-sectional Data

We devise a novel inference algorithm to effectively solve the cancer progression model reconstruction problem. Our empirical analysis of the accuracy and convergence rate of our algorithm, CAncer PRogression Inference (CAPRI), shows that it outperforms the state-of-the-art algorithms addressing similar problems. Motivation: Several cancer-related genomic data have become available (e.g., The Cancer Genome Atlas, TCGA) typically involving hundreds of patients. At present, most of these data are aggregated in a cross-sectional fashion providing all measurements at the time of diagnosis.Our goal is to infer cancer progression models from such data. These models are represented as directed acyclic graphs (DAGs) of collections of selectivity relations, where a mutation in a gene A selects for a later mutation in a gene B. Gaining insight into the structure of such progressions has the potential to improve both the stratification of patients and personalized therapy choices. Results: The CAPRI algorithm relies on a scoring method based on a probabilistic theory developed by Suppes, coupled with bootstrap and maximum likelihood inference. The resulting algorithm is efficient, achieves high accuracy, and has good complexity, also, in terms of convergence properties. CAPRI performs especially well in the presence of noise in the data, and with limited sample sizes. Moreover CAPRI, in contrast to other approaches, robustly reconstructs different types of confluent trajectories despite irregularities in the data.We also report on an ongoing investigation using CAPRI to study atypical Chronic Myeloid Leukemia, in which we uncovered non trivial selectivity relations and exclusivity patterns among key genomic events.

math.ST↗

arXiv2015

On RAF Sets and Autocatalytic Cycles in Random Reaction Networks

The emergence of autocatalytic sets of molecules seems to have played an important role in the origin of life context. Although the possibility to reproduce this emergence in laboratory has received considerable attention, this is still far from being achieved. In order to unravel some key properties enabling the emergence of structures potentially able to sustain their own existence and growth, in this work we investigate the probability to observe them in ensembles of random catalytic reaction networks characterized by different structural properties. From the point of view of network topology, an autocatalytic set have been defined either in term of strongly connected components (SCCs) or as reflexively autocatalytic and food-generated sets (RAFs). We observe that the average level of catalysis differently affects the probability to observe a SCC or a RAF, highlighting the existence of a region where the former can be observed, whereas the latter cannot. This parameter also affects the composition of the RAF, which can be further characterized into linear structures, autocatalysis or SCCs. Interestingly, we show that the different network topology (uniform as opposed to power-law catalysis systems) does not have a significantly divergent impact on SCCs and RAFs appearance, whereas the proportion between cleavages and condensations seems instead to play a role. A major factor that limits the probability of RAF appearance and that may explain some of the difficulties encountered in laboratory seems to be the presence of molecules which can accumulate without being substrate or catalyst of any reaction.

nlin.AO↗

arXiv2014

Inferring tree causal models of cancer progression with probability raising

Existing techniques to reconstruct tree models of progression for accumulative processes, such as cancer, seek to estimate causation by combining correlation and a frequentist notion of temporal priority. In this paper, we define a novel theoretical framework called CAPRESE (CAncer PRogression Extraction with Single Edges) to reconstruct such models based on the notion of probabilistic causation defined by Suppes. We consider a general reconstruction setting complicated by the presence of noise in the data due to biological variation, as well as experimental or measurement errors. To improve tolerance to noise we define and use a shrinkage-like estimator. We prove the correctness of our algorithm by showing asymptotic convergence to the correct tree under mild constraints on the level of noise. Moreover, on synthetic data, we show that our approach outperforms the state-of-the-art, that it is efficient even with a relatively small number of samples and that its performance quickly converges to its asymptote as the number of samples increases. For real cancer datasets obtained with different technologies, we highlight biologically significant differences in the progressions inferred with respect to other competing techniques and we also show how to validate conjectured biological relations with progression models.

math.ST↗

arXiv2014

A stochastic model of catalytic reaction networks in protocells

Protocells are supposed to have played a key role in the self-organizing processes leading to the emergence of life. Existing models either (i) describe protocell architecture and dynamics, given the existence of sets of collectively self-replicating molecules for granted, or (ii) describe the emergence of the aforementioned sets from an ensemble of random molecules in a simple experimental setting (e.g. a closed system or a steady-state flow reactor) that does not properly describe a protocell. In this paper we present a model that goes beyond these limitations by describing the dynamics of sets of replicating molecules within a lipid vesicle. We adopt the simplest possible protocell architecture, by considering a semi-permeable membrane that selects the molecular types that are allowed to enter or exit the protocell and by assuming that the reactions take place in the aqueous phase in the internal compartment. As a first approximation, we ignore the protocell growth and division dynamics. The behavior of catalytic reaction networks is then simulated by means of a stochastic model that accounts for the creation and the extinction of species and reactions. While this is not yet an exhaustive protocell model, it already provides clues regarding some processes that are relevant for understanding the conditions that can enable a population of protocells to undergo evolution and selection.

q-bio.MN↗

arXiv2013

A model of protocell based on the introduction of a semi-permeable membrane in a stochastic model of catalytic reaction networks

In this work we introduce some preliminary analyses on the role of a semi-permeable membrane in the dynamics of a stochastic model of catalytic reaction sets (CRSs) of molecules. The results of the simulations performed on ensembles of randomly generated reaction schemes highlight remarkable differences between this very simple protocell description model and the classical case of the continuous stirred-tank reactor (CSTR). In particular, in the CSTR case, distinct simulations with the same reaction scheme reach the same dynamical equilibrium, whereas, in the protocell case, simulations with identical reaction schemes can reach very different dynamical states, despite starting from the same initial conditions.

cs.CE↗

arXiv2013

Analysis of the spatial and dynamical properties of a multiscale model of intestinal crypts

The preliminary analyses on a multiscale model of intestinal crypt dynamics are here presented. The model combines a morphological model, based on the Cellular Potts Model (CPM), and a gene regulatory network model, based on Noisy Random Boolean Networks (NRBNs). Simulations suggest that the stochastic differentiation process is itself sufficient to ensure the general homeostasis in the asymptotic states, as proven by several measures.

cs.CE↗