One of the most important and challenging ``knowledge extraction' tasks in bioinformatics is the reverse engineering of gene regulatory networks (GRNs) from DNA microarray gene expression data. Indeed, as a result of the development of high-throughput data-collection techniques, biology is experiencing a data flood phenomenon that pushes biologists toward a new view of biology--systems biology--that aims at system-level understanding of biological systems.<p><p>Unfortunately, even for small model organisms such as the yeast Saccharomyces cerevisiae, the number p of genes is much larger than the number n of expression data samples. The dimensionality issue induced by this ``small n, large p' data setting renders standard statistical learning methods inadequate. Restricting the complexity of the models enables to deal with this serious impediment. Indeed, by introducing (a priori undesirable) bias in the model selection procedure, one reduces the variance of the selected model thereby increasing its accuracy.<p><p>Gaussian graphical models (GGMs) have proven to be a very powerful formalism to infer GRNs from expression data. Standard GGM selection techniques can unfortunately not be used in the ``small n, large p' data setting. One way to overcome this issue is to resort to regularization. In particular, shrinkage estimators of the covariance matrix--required to infer GGMs--have proven to be very effective. Our first contribution consists in a new shrinkage estimator that improves upon existing ones through the use of a Monte Carlo (parametric bootstrap) procedure.<p><p>Another approach to GGM selection in the ``small n, large p' data setting consists in reverse engineering limited-order partial correlation graphs (q-partial correlation graphs) to approximate GGMs. Our second contribution consists in an inference algorithm, the q-nested procedure, that builds a sequence of nested q-partial correlation graphs to take advantage of the smaller order graphs' topology to infer higher order graphs. This allows us to significantly speed up the inference of such graphs and to avoid problems related to multiple testing. Consequently, we are able to consider higher order graphs, thereby increasing the accuracy of the inferred graphs.<p><p>Another important challenge in bioinformatics is the prediction of gene function. An example of such a prediction task is the identification of genes that are targets of the nitrogen catabolite repression (NCR) selection mechanism in the yeast Saccharomyces cerevisiae. The study of model organisms such as Saccharomyces cerevisiae is indispensable for the understanding of more complex organisms. Our third contribution consists in extending the standard two-class classification approach by enriching the set of variables and comparing several feature selection techniques and classification algorithms.<p><p>Finally, our fourth contribution formulates the prediction of NCR target genes as a network inference task. We use GGM selection to infer multivariate dependencies between genes, and, starting from a set of genes known to be sensitive to NCR, we classify the remaining genes. We hence avoid problems related to the choice of a negative training set and take advantage of the robustness of GGM selection techniques in the ``small n, large p' data setting. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

Hyaluronic acid (HA) is an important substance, which is mostly used in pharmaceutical and cosmetic industry. This substance is commonly found in the human body. HA is one of the factors contributing to virulence of microorganisms. Some bacterial strains produce hyaluronic acid in the form of a mucoid capsule that encapsulates the cell to protect bacteria against the immune system of the host organism. One of the main producers is the bacterial strain Streptococcus equi subsp. zooepidemicus. Contipro a.s. uses the strain CO4A to produce hyaluronic acid in large scale. The production strain was obtained by random mutagenesis by UV light. The aim of the work was to study changes in the genome, which led to a significant increase in hyaluronic acid production, using DNA microarray and real-time PCR (qPCR). The genome of the strain CO4A was sequenced and compared to reference ATCC35246 [1]. The size of the genome is 2,167,251 bp and 83 relevant variants (59 SNV and 34 indels) have been identified. Variants in coding regions were annotated and amino acid sequence changes were determined. In SNV mutations there was a change in the amino acid sequence in 45 cases. The change was identified in every case of indel mutations. The expression level of selected groups of genes was monitored in both strains by the method of DNA microarrays. A cascade of increased expression level of amino sugar metabolism genes leading to the synthesis of UDP-N-acetyl glucosamine was observed in strain CO4A (the increase in expression level of these genes compared to ATCC35246 was on average 28 %). Subsequently, the expression of selected genes was verified by qPCR. There was no significant difference in the expression level of the has operon genes of both strains. The effect of supplementation of the culture medium with N-acetylglucosamine (GlcNAc), which is one of the precursors of HA synthesis, was also studied by qPCR. A positive effect of the supplementation of the culture medium with external GlcNAc in the CO4A strain has been recorded. Also, the supplementation has positive effect on the yield of HA from the medium (increase in yield was on average by 17 %). GlcNAc has been shown to have a positive effect on the yield of HA in ATCC35246 strain as well (increase in yield was 9 % on average), but no significant changes in the expression levels were found in selected groups of genes in ATCC35246.

Indiana University-Purdue University Indianapolis (IUPUI) / Histone acetylation has been linked to developmental changes in gene expression and is a validated drug target of apicomplexan parasites, but little is known about the roles of individual histone modifying enzymes and how they are recruited to target genes. The protozoan parasite Toxoplasma gondii (phylum Apicomplexa) is unusual among invertebrates in possessing two GCN5-family lysine acetyltransferases (KATs). While GCN5a is required for gene expression in response to alkaline stress, this KAT is dispensable for parasite proliferation in normal culture conditions. In contrast, GCN5b cannot be disrupted, suggesting it is essential for Toxoplasma viability. To further explore the function of GCN5b, we generated clonal parasites expressing an inducible HA-tagged form of GCN5b containing a point mutation that ablates enzymatic activity (E703G). Stabilization of this dominant-negative form of GCN5b was mediated through ligand-binding to a destabilization domain (dd) fused to the protein. Induced accumulation of the ddHAGCN5b(E703G) protein led to a rapid arrest in parasite replication. Growth arrest was accompanied by a decrease in histone H3 acetylation at specific lysine residues as well as reduced expression of GCN5b target genes in GCN5b(E703G) parasites, which were identified using chromatin immunoprecipitation coupled with microarray hybridization (ChIP-chip). We also demonstrate that GCN5b interacts with AP2-domain proteins, which are plant-like transcription factors in Apicomplexa. The interactions between GCN5b, AP2IX-7, and AP2X-8 were confirmed by reciprocal co-immunoprecipitation and revealed a “core complex” that includes the co-activator ADA2-A, TFIID subunits, LEO1 polymerase-associated factor (Paf1) subunit, and RRM proteins. The dominant-negative phenotype of ddHAGCN5b(E703G) parasites, considered with the proteomics and ChIP-chip data, indicate that GCN5b plays a central role in transcriptional and chromatin remodeling complexes. We conclude that GCN5b has a non-redundant and indispensable role in regulating gene expression required during the Toxoplasma lytic cycle.

Στην παρούσα διδακτορική διατριβή προτείνονται μεθοδολογίες υπολογιστικής νοημοσύνης για την επεξεργασία και ανάλυση δεδομένων γονιδιακής έκφρασης μικροσυστοιχιών cDNA. Πιο συγκεκριμένα, στο πρώτο σκέλος αναπτύσσονται δύο νέες προσεγγίσεις για την εύρωστη εκτίμηση και διόρθωση του θορύβου υποβάθρου: η διόρθωση υποβάθρου βάσει εκατοστημορίων και η διόρθωση υποβάθρου βάσει παλινδρόμησης loess. Οι προσεγγίσεις αυτές καινοτομούν κυρίως στο ότι χρησιμοποιούν μία εύρωστη εκτίμηση του θορύβου υποβάθρου, γεγονός που τις καθιστά ιδανικές σε περιπτώσεις, όπου τα δεδομένα είναι θορυβώδη. Επιπροσθέτως, αναπτύσσεται ένα νέο, γενικής χρήσεως, πλαίσιο για τη συστηματική αξιολόγηση του βαθμού επίδρασης των μεθόδων διόρθωσης υποβάθρου. Μέσω του πλαισίου αυτού, οι δύο προτεινόμενες προσεγγίσεις, καθώς και άλλες ευρέως χρησιμοποιούμενες μέθοδοι, αξιολογούνται βάσει εφαρμογής τους σε διαφορετικά σύνολα δεδομένων αυτο-υβριδοποίησης, με τις πρώτες να εμφανίζουν ιδιαιτέρως καλή απόδοση. Το πλαίσιο αυτό καινοτομεί στο ότι ενσωματώνει νέα κριτήρια και τρόπους γραφικής απεικόνισης. Τόσο οι προτεινόμενες μέθοδοι εκτίμησης και διόρθωσης θορύβου υποβάθρου, όσο και το πλαίσιο συστηματικής αξιολόγησής τους, συνιστούν μία νέα, ενδελεχή μελέτη που προσανατολίζει στην εφαρμογή ή απόρριψη μίας συγκεκριμένης προσέγγισης, συνεισφέροντας εν τέλει στην κατάκτηση καλλίτερης ποιότητας δεδομένων μικροσυστοιχιών. Επίσης, στο δεύτερο σκέλος της διατριβής αναπτύσσεται ένα νέο, ολοκληρωμένο και γενικής χρήσεως πλαίσιο ανάλυσης δεδομένων μικροσυστοιχιών ούτως, ώστε να διερευνηθεί το ζήτημα εάν στην T-λευχαιμική κυτταρική σειρά CCRF-CEM επικρατούν εγγενείς ή επίκτητοι μηχανισμοί αντοχής στην πρεδνιζολόνη. Συγκεκριμένα, καταλλήλως επιλεχθέντα δεδομένα μικροσυστοιχιών cDNA – που διευκολύνουν την εξέταση τόσο της εξαρτώμενης από τη συγκέντρωση δράσης, όσο και της δυναμικής της ανταπόκρισης στην πρεδνιζολόνη (πρώιμη και όψιμη δράση) – γίνονται αντικείμενο επεξεργασίας και ενδελεχούς ανάλυσης, και βάσει συγκεκριμένων, προ-διατυπωμένων συλλογισμών, προσεγγίζεται το εν λόγω ερώτημα. Το πλαίσιο αυτό είναι καινοτόμο, εφόσον, πέραν του ότι ενσωματώνει μία πρωτότυπη ακολουθία μεθόδων, προσεγγίζει συστηματικά το πρόβλημα της εγγενούς ή επίκτητης αντοχής, συνεισφέροντας, έτσι, στην ευρύτερη προσπάθεια διερεύνησης των επακριβών μηχανισμών αντοχής των λευχαιμικών κυττάρων στα γλυκοκορτικοειδή. Τα αποτελέσματα από την εφαρμογή του στα δεδομένα της εν λόγω κυτταρικής σειράς συνηγορούν υπέρ της ύπαρξης μίας σύνθετης ανταπόκρισης του υπό μελέτη συστήματος στα γλυκοκορτικοειδή, η οποία όμως τείνει περισσότερο προς έναν εγγενή μηχανισμό αντοχής. / In the present Ph.D. thesis, computational intelligence methods for processing and analyzing cDNA microarray gene expression data are designed and developed. More specifically, in the first part of this thesis, the problem of background estimation and correction of two-channel microarray data is addressed and two novel algorithms are proposed, namely the percentiles-based and the loess-based background correction methods. Both approaches are based on the multiplicative model of background, while utilizing robust background noise estimators, thus making them ideal for noisy datasets. Furthermore, a new, generic framework for the systematic evaluation of the impact of the background estimating methodologies is suggested, whereupon the aforementioned methods as well as other approaches are evaluated by application to various publicly available self-self hybridization datasets. As suggested by this thorough, comparative evaluation our algorithms perform very well regarding noise reduction. The evaluation framework, which is based mainly on different and widely used statistical measures, incorporates new criteria and visualization methods. Moreover, it represents a novel, detailed contribution to the examination of the impact of background correction methods to the final interpretation of microarray experiments, conferring explicit guidance on the pros and cons of them and when they should be applied. Additionally, in the second part of this thesis, a new, generic, computational microarray data analysis framework is described, in order to examine the hypothesis of whether the resistant T-cell leukemia cell line CCRF-CEM posses an intrinsic or exert an acquired mechanism of resistance and to investigate the molecular imprint of this, upon prednisolone treatment. More analytically, using the above explained computational analysis workflow, microarray data that enable the examination of both the dose effect of prednisolone exposure and the dynamics (early and late) of the molecular response of the cells at the transcriptomic layer, are systematically analyzed based on specific, predefined formulations. The analysis of the results supports a complex mechanism of action for the cells which seems to favor though more the intrinsic mechanism of resistance.

This thesis addresses the use of machine learning techniques to develop clinical diagnostic tools for breast cancer using molecular data. These tools are designed to assist physicians in their evaluation of the clinical outcome of breast cancer (referred to as prognosis).<p>The traditional approach to evaluating breast cancer prognosis is based on the assessment of clinico-pathologic factors known to be associated with breast cancer survival. These factors are used to make recommendations about whether further treatment is required after the removal of a tumor by surgery. Treatment such as chemotherapy depends on the estimation of patients' risk of relapse. Although current approaches do provide good prognostic assessment of breast cancer survival, clinicians are aware that there is still room for improvement in the accuracy of their prognostic estimations.<p>In the late nineties, new high throughput technologies such as the gene expression profiling through microarray technology emerged. Microarrays allowed scientists to analyze for the first time the expression of the whole human genome ("transcriptome"). It was hoped that the analysis of genome-wide molecular data would bring new insights into the critical, underlying biological mechanisms involved in breast cancer progression, as well as significantly improve prognostic prediction. However, the analysis of microarray data is a difficult task due to their intrinsic characteristics: (i) thousands of gene expressions are measured for only few samples; (ii) the measurements are usually "noisy"; and (iii) they are highly correlated due to gene co-expressions. Since traditional statistical methods were not adapted to these settings, machine learning methods were picked up as good candidates to overcome these difficulties. However, applying machine learning methods for microarray analysis involves numerous steps, and the results are prone to overfitting. Several authors have highlighted the major pitfalls of this process in the early publications, shedding new light on the promising but overoptimistic results. <p>Since 2002, large comparative studies have been conducted in order to identify the key characteristics of successful methods for class discovery and classification. Yet methods able to identify robust molecular signatures that can predict breast cancer prognosis have been lacking. To fill this important gap, this thesis presents an original methodology dealing specifically with the analysis of microarray and survival data in order to build prognostic models and provide an honest estimation of their performance. The approach used for signature extraction consists of a set of original methods for feature transformation, feature selection and prediction model building. A novel statistical framework is presented for performance assessment and comparison of risk prediction models.<p>In terms of applications, we show that these methods, used in combination with a priori biological knowledge of breast cancer and numerous public microarray datasets, have resulted in some important discoveries. In particular, the research presented here develops (i) a robust model for the identification of breast molecular subtypes and (ii) a new prognostic model that takes into account the molecular heterogeneity of breast cancers observed previously, in order to improve traditional clinical guidelines and state-of-the-art gene signatures./Cette thèse concerne le développement de techniques d'apprentissage (machine learning) afin de mettre au point de nouveaux outils cliniques basés sur des données moleculaires. Nous avons focalisé notre recherche sur le cancer du sein, un des cancers les plus fréquemment diagnostiqués. Ces outils sont développés dans le but d'aider les médecins dans leur évaluation du devenir clinique des patients cancéreux (cf. le pronostique).<p>Les approches traditionnelles d'évaluation du pronostique d'un patient cancéreux se base sur des critères clinico-pathologiques connus pour être prédictifs de la survie. Cette évaluation permet aux médecins de décider si un traitement est nécessaire après l'extraction de la tumeur. Bien que les outils d'évaluation traditionnels sont d'une aide importante, les cliniciens sont conscients de la nécessité d'améliorer de tels outils.<p>Dans les années 90, de nouvelles technologies à haut-débit, telles que le profilage de l'expression génique par biopuces à ADN (microarrays), ont été mises au point afin de permettre aux scientifiques d'analyser l'expression de l'entièreté du génôme de cellules cancéreuses. Ce nouveau type de données moléculaires porte l'espoir d'améliorer les outils pronostiques traditionnels et d'approfondir nos connaissances concernant la génèse du cancer du sein. Cependant ces données sont extrêmement difficiles à analyser à cause (i) de leur haute dimensionalité (plusieurs dizaines de milliers de gènes pour seulement quelques centaines d'expériences); (ii) du bruit important dans les mesures; (iii) de la collinéarité entre les mesures dûe à la co-expression des gènes.<p>Depuis 2002, des études comparatives à grande échelle ont permis d'identifier les méthodes performantes pour l'analyse de groupements et la classification de données microarray, négligeant l'analyse de survie pertinente pour le pronostique dans le cancer du sein. Pour pallier ce manque, cette thèse présente une méthodologie originale adaptée à l'analyse de données microarray et de survie afin de construire des modèles pronostiques performants et robustes. <p>En termes d'applications, nous montrons que cette méthodologie, utilisée en combinaison avec des connaissances biologiques a priori et de nombreux ensembles de données publiques, a permis d'importantes découvertes. En particulier, il résulte de la recherche presentée dans cette thèse, le développement d'un modèle robuste d'identification des sous-types moléculaires du cancer du sein et de plusieurs signatures géniques améliorant significativement l'état de l'art au niveau pronostique. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

Les biopuces à ADN permettent d’étudier à une échelle génomique une très grande variété de questions sur la physiologie et la différenciation cellulaires. Elles ont ainsi contribué de manière considérable aux progrès récents de nombreux domaines de la biologie et occuperont bientôt une place de choix dans le secteur du diagnostic médical. C’est la levure Saccharomyces cerevisiae qui a servi de modèle pour le développement de la première biopuce génomique. L’application de cette approche à la levure a permis d’explorer sous un angle nouveau l’étude de ses différents états de différenciation, de son cycle cellulaire, et de sa capacité d’adaptation à diverses conditions nutritionnelles ou à des conditions environnementales induisant un stress cellulaire. Plusieurs études ont plus particulièrement examiné la réponse des cellules de levure à une carence en azote ou en acides aminés. Cependant, une étude systématique de la réponse transcriptionnelle de la cellule aux différentes sources d’azote n’a jamais été entreprise en croissance confinée à l'état de régime. S. cerevisiae est capable d’utiliser plus d’une vingtaine de substances en tant que sources uniques d’azote pour la croissance. On distingue parmi les sources d’azote celles qui permettent une croissance optimale, appelées « bonnes » sources d’azote, des autres, appelées « mauvaises » sources d’azote. La levure possède plusieurs systèmes de régulation lui permettant de s'adapter à la condition azotée. Au niveau transcriptionnel, on recense trois régulations générales – la NCR (répression catabolique azotée), le GAAC (le contrôle général de la biosynthèse des acides aminés) et le système SPS (Ssy1-Ptr3-Ssy5) – et une multitude de régulations plus spécifiques.<p>En utilisant la technique des puces à ADN, nous avons généré une matrice d'expression de l'ensemble des gènes de la levure en croissance confinée à l'état de régime dans un milieu de culture contenant une parmi 21 sources d'azote différentes. Nous avons pu ainsi recenser systématiquement 506 gènes soumis à une régulation transcriptionnelle dépendante de l'azote.<p>En nous basant sur ces résultats, nous avons pu décrire l'ensemble des régulations transcriptionnelles engagées dans l'adaptation à la source d'azote fournie dans le milieu de culture. Parallèlement, nous avons défini deux grands groupes de sources d'azote en fonction du transcriptome de S. cerevisiae. Le premier groupe rassemble les composés qui exercent une répression catabolique azotée forte et dont la liste a été complétée. Fait nouveau, nous montrons que ces mêmes composés enclenchent aussi l'activation de la réponse aux protéines mal repliées (UPR). Au contraire, lorsque la source d'azote appartient au second groupe que nous avons défini, non seulement la croissance des levures est plus lente, la NCR levée et la réponse aux protéines mal repliées réprimée, mais nous montrons de façon inattendue que le contrôle général de la biosynthèse des acides aminés est activé. Plusieurs autres régulations qui ne sont pas impliquées dans le métabolisme azoté présentent un comportement différent en fonction de la source d'azote fournie. C'est le cas notamment des gènes dont l’expression est régulée selon l’apport en zinc et qui sont moins exprimés sur le milieu urée. De même, les gènes impliqués dans les résistances multiples aux drogues sont activés par le tryptophane.<p>En confrontant nos résultats à ceux obtenus dans le cadre de travaux indépendants, nous avons proposé plus d'une cinquantaine de nouveaux gènes cibles de la NCR. Beaucoup d'entre eux n'ont jamais été caractérisés expérimentalement. En utilisant des techniques avancées d'analyse de séquences primaires de protéines, nous avons pu proposer une fonction pour plusieurs de ces gènes. Ces analyses bioinformatiques et la réalisation d’expériences complémentaires à l’aide de biopuces à ADN nous ont permis de proposer que l'un d'entre eux code pour une protéine impliquée dans la déstabilisation d'ARN messagers lors de la carence azotée. Nous avons aussi identifié plusieurs nouveaux gènes appartenant à des régulons spécifiquement activés en réponse à un nombre restreint de sources d'azote. Il est probable que ces gènes soient impliqués dans le catabolisme des sources d'azote sur lesquelles ils sont activés. / Doctorat en sciences, Spécialisation biologie moléculaire / info:eu-repo/semantics/nonPublished

Indiana University-Purdue University Indianapolis (IUPUI) / During pregnancy, maternal liver exhibits robust growth to meet the metabolic demands of the developing placenta and fetus. Although hepatocyte hypertrophy and hyperplasia are seen in the maternal liver, the molecular and cellular mechanisms mediating the maternal hepatic adaptations to pregnancy is poorly understood. Previous microarray analysis revealed a most upregulated gene named Ascl1, a transcription factor essential for neural development, in the maternal liver at mid-gestation. The aims of the study were to (1) validate the activation of Ascl1 gene; (2) identify Ascl1-expressing cells; and (3) determine the fate of Ascl1-expressing cells, in the maternal liver during the course of gestation. Timed pregnancy was setup in mice and the maternal livers were collected at various stages of gestation. Maternal hepatic Ascl1 mRNA expression was evaluated by qRT-PCR and northern blotting. The results demonstrated that the transcript level of maternal hepatic Ascl1 is exponentially increased during the second half of pregnancy in comparison with a non-pregnant state. Using a Ascl1-GFP mouse model generated by others to monitor the behavior of neural progenitor cells, we found that maternal hepatic Ascl1-expressing cells are non-parenchymal cells, very small in size, and expanding during pregnancy. To map the fate of this cell population, we generated an in vivo tracing mouse model named Ascl1-CreERT2/ROSA26-LacZ. Using this model, we permanently labeled maternal hepatic Ascl1-expressing cells at midgestation by giving tamoxifen and analyzed the labeled cells in the maternal liver prior to parturition. We observed that the initial small Ascl1-expressing cells undergoing expansion at mid-gestation eventually became hepatocyte-like cells at the end stage of pregnancy. Taken together, our findings strongly suggest that Ascl1-expressing cells represent a novel population of hepatic progenitor cells and they can differentiate along hepatocyte lineage and contribute to pregnancy-induced maternal liver growth. Further studies are needed to firmly establish the nature and property of maternal hepatic Ascl1-expressing cells. At this stage, we have gained significant insights into the cellular mechanism by which the maternal liver adapts to pregnancy.

Indiana University-Purdue University Indianapolis (IUPUI) / Phormia regina (Meigen), commonly known as the black blow fly is a dipteran that belongs to the family Calliphoridae. Calliphorids play an important role in various research fields including ecology, medical studies, veterinary and forensic sciences. P. regina, a non-model organism, is one of the most common forensically relevant insects in North America and is typically used to assist in estimating postmortem intervals (PMI). To better understand the roles P. regina plays in the numerous research fields, we re-constructed its genome using next generation sequencing technologies. The focus was on generating a reference genome through de novo assembly of high-throughput short read sequences. Following assembly, genetic markers were identified in the form of microsatellites and single nucleotide polymorphisms (SNPs) to aid in future population genetic surveys of P. regina. A total 530 million 100 bp paired-end reads were obtained from five pooled male and female P. regina flies using the Illumina HiSeq2000 sequencing platform. A 524 Mbp draft genome was assembled using both sexes with 11,037 predicted genes. The draft reference genome assembled from this study provides an important resource for investigating the genetic diversity that exists between and among blow fly species; and empowers the understanding of their genetic basis in terms of adaptations, population structure and evolution. The genomic tools will facilitate the analysis of genome-wide studies using modern genomic techniques to boost a refined understanding of the evolutionary processes underlying genomic evolution between blow flies and other insect species.

Indiana University-Purdue University Indianapolis (IUPUI) / Break-induced replication (BIR) is a mechanism to repair double-strand breaks (DSBs) that possess only a single end that can find homology in the genome. This situation can result from the collapse of replication forks or telomere erosion. BIR frequently produces various genetic instabilities including mutations, loss of heterozygosity, deletions, duplications, and template switching that can result in copy-number variations (CNVs). An important type of genomic rearrangement specifically linked to BIR is half crossovers (HCs), which result from fusions between parts of recombining chromosomes. Because HC formation produces a fused molecule as well as a broken chromosome fragment, these events could be highly destabilizing. Here I demonstrate that HC formation results from the interruption of BIR caused by a defective replisome or premature onset of mitosis. Additionally, I document the existence of half crossover instability cascades (HCC) that resemble cycles of non-reciprocal translocations (NRTs) previously described in human tumors. I postulate that HCs represent a potent source of genetic destabilization with significant consequences that mimic those observed in human diseases, including cancer.