Computing nearest neighbour interchange distances between ranked phylogenetic trees

Phylogenetic tree-search is a major aspect of many evolutionary studies. Several tree rearrangement algorithms are available for tree-search, but it is hard to draw general conclusions about their relative performance because many effects are data set specific and can be highly dependent on individual implementations (e.g., RAxML or phyml). Using only the structure of the rearrangements proposed by the Nearest Neighbor Interchange (NNI) algorithm, we show tree-search can prematurely terminate if it encounters multifurcating trees. We validate the relevance of this result by demonstrating that in real data the majority of possible bifurcating trees potentially encountered during tree-search are actually multifurcations, which suggests NNI would be expected to perform poorly. We also show that the star-decomposition algorithm is a special case of two other popular tree-search algorithms, subtree pruning and regrafting (SPR) and tree bisection and reconnection (TBR), which means that these two algorithms can efficiently escape when they encounter multifurcations. We caution against the use of the NNI algorithm and for most applications we recommend the use of more robust tree-search algorithms, such as SPR and TBR.

A note on the relaxation time of two Markov chains on rooted phylogenetic tree spaces

In phylogenomic analysis the collection of trees with identical score (maximum likelihood or parsimony score) may hamper tree search algorithms. Such collections are coined phylogenetic terraces. For sparse supermatrices with a lot of missing data, the number of terraces and the number of trees on the terraces can be very large. If terraces are not taken into account, a lot of computation time might be unnecessarily spent to evaluate many trees that in fact have identical score. To save computation time during the tree search, it is worthwhile to quickly identify such cases. The score of a species tree is the sum of scores for all the so-called induced partition trees. Therefore, if the topological rearrangement applied to a species tree does not change the induced partition trees, the score of these partition trees is unchanged. Here, we provide the conditions under which the three most widely used topological rearrangements (nearest neighbor interchange, subtree pruning and regrafting, and tree bisection and reconnection) change the topologies of induced partition trees. During the tree search, these conditions allow us to quickly identify whether we can save computation time on the evaluation of newly encountered trees. We also introduce the concept of partial terraces and demonstrate that they occur more frequently than the original “full” terrace. Hence, partial terrace is the more important factor of timesaving compared to full terrace. Therefore, taking into account the above conditions and the partial terrace concept will help to speed up the tree search in phylogenomic inference.

A central theme in phylogenetics is the reconstruction and analysis of evolutionary trees from a given set of data. To determine the optimal search methods for reconstructing trees, it is crucial to understand the size and structure of the neighborhoods of trees under tree rearrangement operations. The diameter and size of the immediate neighborhood of a tree have been well-studied; however, little is known about the number of trees at distance two, three, or (more generally) $k$ from a given tree. In this paper we provide a number of exact and asymptotic results concerning these quantities and identify some key aspects of tree shape that play a role in determining these quantities. We obtain several new results for two of the main tree rearrangement operations---nearest neighbor interchange and subtree prune and regraft---as well as for the Robinson--Foulds metric on trees.

The Minimum Evolution (ME) approach to phylogeny estimation has been shown to be statistically consistent when it is used in conjunction with ordinary least-squares (OLS) fitting of a metric to a tree structure. The traditional approach to using ME has been to start with the Neighbor Joining (NJ) topology for a given matrix and then do a topological search from that starting point. The first stage requires O(n(3)) time, where n is the number of taxa, while the current implementations of the second are in O(p n(3)) or more, where p is the number of swaps performed by the program. In this paper, we examine a greedy approach to minimum evolution which produces a starting topology in O(n(2)) time. Moreover, we provide an algorithm that searches for the best topology using nearest neighbor interchanges (NNIs), where the cost of doing p NNIs is O(n(2) + p n), i.e., O(n(2)) in practice because p is always much smaller than n. The Greedy Minimum Evolution (GME) algorithm, when used in combination with NNIs, produces trees which are fairly close to NJ trees in terms of topological accuracy. We also examine ME under a balanced weighting scheme, where sibling subtrees have equal weight, as opposed to the standard "unweighted" OLS, where all taxa have the same weight so that the weight of a subtree is equal to the number of its taxa. The balanced minimum evolution scheme (BME) runs slower than the OLS version, requiring O(n(2) x diam(T)) operations to build the starting tree and O(p n x diam(T)) to perform the NNIs, where diam(T) is the topological diameter of the output tree. In the usual Yule-Harding distribution on phylogenetic trees, the diameter expectation is in log(n), so our algorithms are in practice faster that NJ. Moreover, this BME scheme yields a very significant improvement over NJ and other distance-based algorithms, especially with large trees, in terms of topological accuracy.

The phylogeny reconstruction problem is a fundamental problem in computational molecular biology and biochemical physics. Since the number of data sets has grown substantially in recent years, the accuracy and speed of constructing phylogenies become increasingly critical. Numerous studies have demonstrated that the maximum likelihood (ML) method is the most effective method for reconstructing a phylogenetic tree from sequence data. Conversely, tree bisection and reconnection (TBR) is a tree topology rearrangement method that can generate an extensive tree space. In this paper, we propose an enhanced method for reconstructing phylogenetic trees in which the TBR operation is modified and combined with the minimum evolution principle to filter out some unnecessary reconnected positions to reduce the search time. The experiment results demonstrate that the proposed method can assist other algorithms in constructing more accurate trees within a reasonable time.

Heuristics for unrooted, unranked, and ranked anomaly zones under birth-death models

Finding the best phylogenetic tree under the maximum parsimony optimality criterion is computationally difficult. We quantify the occurrence of such optima for well-behaved sets of data. When nearest neighbor interchange operations are used, multiple local optima can occur even for "perfect" sequence data, which results in hill-climbing searches that never reach a global optimum. In contrast, we show that when neighbors are defined via the subtree prune and regraft metric, there is a single local optimum for perfect sequence data, and thus, every such search finds a global optimum quickly. We further characterize conditions for which sequences simulated under the Cavender-Farris-Neyman and Jukes-Cantor models of evolution yield well-behaved search spaces.

The Maximum Parsimony (MP) problem aims at reconstructing a phylogenetic tree from DNA sequences while minimizing the number of genetic transformations. To solve this NP-complete problem, heuristic methods have been developed, often based on local search. In this article, we focus on the influence of the neighborhood relations. After analyzing the advantages and drawbacks of the well-known Nearest Neighbor Interchange (NNI), Subtree Pruning Regrafting (SPR) and Tree-Bisection-Reconnection (TBR) neighborhoods, we introduce the concept of Progressive Neighborhood (PN) which consists in constraining progressively the size of the neighborhood as the search advances. We empirically show that applied to the Maximum Parsimony problem, this progressive neighborhood turns out to be more efficient and robust than the classic neighborhoods using a descent algorithm. Indeed, it allows to find better solutions with a smaller number of iterations or trees evaluated.

Co-evolution has an important function in the evolution of species and it is clearly manifested in certain scenarios such as host–parasite and predator–prey interactions, symbiosis and mutualism. The extrapolation of the concepts and methodologies developed for the study of species co-evolution at the molecular level has prompted the development of a variety of computational methods able to predict protein interactions through the characteristics of co-evolution. Particularly successful have been those methods that predict interactions at the genomic level based on the detection of pairs of protein families with similar evolutionary histories (similarity of phylogenetic trees: mirrortree). Future advances in this field will require a better understanding of the molecular basis of the co-evolution of protein families. Thus, it will be important to decipher the molecular mechanisms underlying the similarity observed in phylogenetic trees of interacting proteins, distinguishing direct specific molecular interactions from other general functional constraints. In particular, it will be important to separate the effects of physical interactions within protein complexes (‘co-adaptation') from other forces that, in a less specific way, can also create general patterns of co-evolution.

The American Association for Cancer Research (AACR) is the first and largest cancer research organization dedicated to accelerating the conquest of cancer. The stated purpose of the organization when it was founded 115 years ago was “to further the investigation and spread the knowledge of cancer.” Since that time, the AACR has been the driving force to eradicate cancer. Publication of research findings has always been critical to this endeavor, with the 11 founders recognizing the importance of “sharing observations” and the need for a journal to “collect under one cover such contributions as bear in any way upon the general problems of oncology.”The AACR launched its first peer-reviewed scientific journal, The Journal of Cancer Research, in 1916. At the time, it was the only English-language cancer journal in the world. Since then, the AACR portfolio of journals has grown to 10. These journals cover the full spectrum of cancer science and medicine and together, over the years, have published over 100,000 articles and over 130,000 meeting abstracts. This is an immense contribution to the body of knowledge on cancer and has ignited communication among cancer scientists and physicians, helping to catalyze a revolution in the field.We are honored to serve as the current Editors-in-Chief of the AACR journals. Each of the 10 journals was launched by the AACR to meet the scientific information needs of its membership and all those engaged in cancer research. Under our leadership, the journals engage closely and collaboratively with the community to facilitate the presentation and wide dissemination of research findings. Importantly, the AACR is a not-for-profit publisher. This means that any surplus from the AACR journals program is used to help support the cancer research community and the mission of the AACR to prevent and cure all cancers by contributing to funding for AACR grants, scientific conferences, and science policy and advocacy efforts. Here, we highlight the unique roles of our journals within the family of AACR journals and how they serve to advance the mission of the AACR.At the inception of the AACR, there was a clear need for journals focused on cancer research to assemble key studies to help guide the field. Cancer Research was launched in 1941 as the flagship journal of the AACR from a rich ancestry that started with its predecessors: The Journal of Cancer Research (1916–1930) and The American Journal of Cancer Research (1931–1940). From the very beginning, Cancer Research has published landmark papers (https://aacrjournals.org/cancerres/pages/landmark_articles) that have enriched our knowledge of cancer biology and treatment and have provided the foundation upon which major breakthroughs have been built. In the first year of publication, Cancer Research featured seminal studies on the effect of castration on prostate cancer and the mechanism of carcinogenesis that shaped the field (https://www.aacr.org/wp-content/uploads/2019/11/CancerResearch_75Anniversary_1941-2016.pdf). Cancer Research has continued to publish articles that cover the full range of research, spanning from investigations on drug discovery, cancer molecular mechanisms, and the tumor microenvironment to clinical and population studies. As cancer research has evolved over the past century, new AACR journals have emerged to embrace and support expanding fields, providing specialized venues for studies that deepen our understanding of areas that were historically covered by Cancer Research. In this regard, the “pedigree” of AACR journals stemmed from the principles set by Cancer Research, which remains a venue for publishing fundamental studies in all areas of cancer research, including emerging areas such as cancer data science and mathematics. As a society journal, Cancer Research serves the community through its editorial process that is guided by the judgment of international academic editors who provide rich expertise in various areas as practicing researchers. Further, the Journal embraces early career investigators and has recognized these researchers with the prestigious Cancer Research Early Career Award. Guided by its extensive history, Cancer Research hopes to motivate and support the next generation of scientists to lessen the burden of cancer through their innovative research.As cancer science and medicine entered the molecular biology era, it became clear that a communication channel was needed to emphasize basic science studies in oncology—studies that report significant mechanistic findings at the molecular or cellular level. Originally published under the title Cell Growth & Differentiation from 1990 to 2002, Molecular Cancer Research (MCR) was launched as the AACR was beginning to build its now-renowned focus on molecular biology and genetics. The Journal served to help emphasize this focus and attract more scientists into the AACR's membership. As part of this mission, MCR features studies exploring the molecular underpinnings of each of the hallmarks of cancer, with a particular emphasis on studies elucidating oncogenic alterations in metabolism. Cancer metabolism is increasingly recognized as a key vulnerability of tumor biology. MCR is committed to recruiting and disseminating studies featuring novel metabolomic analyses, mass spectrometry, metabolic imaging and tracing, and other approaches that clarify links between metabolism and cancer, with the goal of identifying metabolic vulnerabilities and adding to the armamentarium of anticancer interventions. In addition to publishing impactful basic cancer research studies, the Journal aims to promote diversity of thought and perspective within the basic cancer research community. To achieve this goal, MCR has launched an initiative in which senior investigators are paired with rising stars in their fields to write and edit review articles jointly. This initiative encourages collaboration among researchers by integrating historical perspectives with new directions to provide novel, actionable insights for the field.Cancer Epidemiology, Biomarkers & Prevention (CEBP) is the leading subspecialty journal for fundamental and applied population science research describing the burden of cancer; uncovering possible causes of cancer and its progression; and informing and evaluating strategies for cancer prevention, early detection, cancer survivorship, and closing the cancer disparities gap. CEBP is unique among AACR journals in focusing on populations at risk for cancer, with cancer, and surviving cancer, especially U.S. and global populations who experience health inequities. Among its many contributions, the Journal has advanced the field of molecular epidemiology of cancer by providing expert peer review and a highly regarded publications forum and has helped launch the careers of many investigators who conduct population science research on cancer. CEBP plays an active role in informing public health practice, policy, and recommendations by publishing articles that are cited as part of the evidence base, including by the FDA, the Centers for Disease Control and Prevention, the U.S. Preventive Services Task Force, and the World Health Organization. Recognizing the societal need to enhance the translation of population science evidence for cancer prevention and control today and in the future, the Journal recently expanded its scope to include cancer care delivery and implementation science research. Like the other AACR journals, CEBP serves the research community well through its editorial rigor and publication of innovative and impactful research addressing contemporary and emerging cancer problems.As biologic insights into cancer rapidly accelerated in the decades after the discovery of the structure of DNA and the armamenta­rium of anticancer agents began to expand, the AACR identified the need for a communication outlet focused on the clinical applications of these discoveries; Clinical Cancer Research was launched to provide this much-needed outlet. Early therapeutic development has since transitioned from rigidly divided trial phases into an adaptive process that encompasses investigation of safety, pharmacokinetics, proof of mechanism, and first evidence of clinical activity. These same trials now frequently serve as the basis for first regulatory approval. The Journal bridges the transition from late preclinical development, when a new therapy is being paired with insights into the molecular features that predict responsiveness to therapy, to clinical validation of those hypotheses, and the Journal has evolved into a premier venue for phase I/II trials. As the only AACR journal with clinical research of all types, including clinical trials, as its sole purview, Clinical Cancer Research has taken the charge of being the journal that accounts for what happens to the discoveries described in basic and translational cancer research when they are tested prospectively for their potential impact on the outcomes of patients with cancer. Since its launch, Clinical Cancer Research has also played a distinct educational role among clinical oncology journals, with a robust collection of review articles to benefit basic scientists with clinical interests and clinical investigators studying new agents with unique biologic mechanisms of action. Clinical Cancer Research has also become an important forum for critical discussions in clinical oncology. Field-leading experts are invited to deliberate on the progress made in a given therapeutic area, focusing on the lessons to be learned from recent successes and failures. Authors from the FDA provide detailed insights into the data that drive regulatory decision-making for each successful therapy. Clinical Cancer Research continues to demonstrate the AACR's long-standing commitment to accelerating research that leads to potential interventions for prevention, early detection, and treatment of cancer and particularly to showcasing clinical trials.Improving and expanding the arsenal of anticancer therapies is an essential goal in our shared mission to prevent and cure all cancers. From before the humble beginnings of nitrogen mustards and vinca alkaloids to the advent of immuno-oncology and rationally designed targeted therapy, the last 115 years have seen quantum leaps in therapeutic strategies and clinical options to combat cancer. Molecular Cancer Therapeutics has been at the forefront of this effort for more than 20 years, publishing top-tier science in the design, synthesis, discovery, and preclinical study of novel therapeutic agents for the treatment and prevention of cancer. The Journal's scope covers anticancer therapeutics broadly, including precision medicine therapies of all types: cell therapies, gene therapies, RNA therapeutics, oncolytic viruses, and vaccines in addition to established regimens such as mole­cular and biologic agents, chemotherapy, and radiotherapy. Among these efforts, the Journal provides a specialized venue for the first disclosure of experimental therapeutics advancing toward clinical trials, providing a platform with international reach to authors. The crucial role of Molecular Cancer Therapeutics among the AACR journals—and, indeed, among all oncology journals—cannot be overstated. The Journal's laser focus on disseminating rigorously vetted research with near-term clinical benefit has enormous potential to impact the entire cancer research community and to improve, extend, and defend the lives of patients with cancer. To that end, Molecular Cancer Therapeutics continues to provide a vital forum for the communication of important discoveries in cancer therapeutics research to the global research community.Effective global cancer control must emphasize efforts to reduce cancer incidence rather than concentrating solely on cancer treatment. To help catalyze such efforts, the AACR launched Cancer Prevention Research (CaPR) in 2008 to be devoted excl­usively to cancer prevention research. From its inception, CaPR has endeavored to be a primary resource for original articles, scholarly reviews, and timely perspectives regarding basic, translational, clinical, and population science investigations related to cancer risk reduction. Five years ago, CaPR published the landmark perspective, “Transforming Cancer Prevention through Precision Medicine and Immune-oncology,” which proposed a Pre-Cancer Atlas and recognized unprecedented opportunities to interrogate the biology of premalignancy (1). The subsequent development of the Pre-Cancer Atlas, together with other advances, represents a “moonshot” opportunity for the field of cancer prevention. CaPR will continue to cultivate high-caliber content from pioneers in fields such as carcinogenesis, the biology of premalignancy, cancer risk assessment, screening, and policy implementation that will transform the way we detect and, ultimately, intercept and prevent cancer. The Journal's mission to reflect the current state of cancer prevention and to catalyze the development of the field supports the AACR mission to prevent and cure all cancers. The community of researchers, educators, advocates, and funders that works together under the umbrella of the AACR and its family of journals is a critical part of the cancer prevention ecosystem. Together, we are publishing research that defines the field of cancer prevention, building a translational bridge between bench and community, and developing the next generation of cancer prevention researchers.In 2011, recognizing the cancer research community's need for a journal that published the most impactful research while placing a premium on streamlined editorial processes and service to authors, the AACR launched Cancer Discovery. Bringing together basic scientists with detailed knowledge of cancer biology and physicians at the vanguard of science-driven clinical trial design, Cancer Discovery immediately changed the landscape of scientific publishing in the field, with many other journals now trying to emulate the unique mix of basic, translational, and clinical research that the Journal pioneered. Cancer Discovery has been at the forefront of the most important developments in cancer over the past decade, from targeted therapy and immunotherapy to technological advances in single-cell sequencing and liquid biopsy that have fundamentally changed our understanding of cancer biology and response to treatment. Preclinical and clinical studies published in Cancer Discovery have laid the foundation for numerous FDA approvals within a short period, and influential commentaries and reviews have galvanized the field and guided policy decisions. The unique editorial team consisting of leading researchers and experienced professional editors works collaboratively with the community to quickly publish and disseminate the strongest, immediately impactful science to support the AACR mission to prevent and cure all cancers.The first cancer immunotherapeutic is famously considered to have been administered in 1891 by William B. Coley, one of the 11 founders of the AACR, when he used a mixture of heat-killed bacteria (subsequently known as Coley's toxins) to create an inflammatory response that could treat a patient with sarcoma. However, it took more than 100 years before the field of cancer immunology and immunotherapy research gained meaningful traction among those working in the broader discipline of cancer science and medicine. The rapid expansion of the field began in the early 2000s as a result of increasing evidence linking the immune system with cancer development and demonstrations that the immune system might be harnessed to treat cancer, and it has continued at an even more remarkable pace since the FDA approval of the first immune checkpoint inhibitor, the CTLA4-specific blocking antibody ipilimumab, in 2011. Recognizing the need to further enhance interactions between researchers in the realm of cancer immunology and immunotherapy and those in all other areas of the cancer research community, and to provide a forum for the presentation of advances in the rapidly evolving field of cancer immunology science, the AACR collaborated with the Cancer Research Institute to launch Cancer Immunology Research in 2013. Since its inaugural issue, the Journal has disseminated exciting discoveries and developments in the field of cancer immunology and immunotherapy, and successfully introduced the central principles of immunology to cancer biologists and clinical investigators. As a journal devoted to the science underlying some of the most transformative therapeutics to have entered the clinic for the treatment of cancer in recent years, Cancer Immunology Research has played a unique role in advancing the mission of the AACR.Unprecedented progress in understanding the pathogenesis of hematologic malignancies and in translating these insights into improved clinical outcomes has created a demand for a high-profile outlet for publishing these discoveries. To address this need and to emphasize the AACR's commitment to the prevention and cure of all types of cancer, Blood Cancer Discovery was launched in 2020 to inspire, facilitate, and broadly disseminate important discoveries about hematologic malignancies. The Journal's model of operation replicates Cancer Discovery's success in creating synergies between the expertise of leaders in the field and the dedication of in-house editorial and publishing teams, enabling rigorous, rapid, and transparent editorial processes. Blood Cancer Discovery has published cutting-edge research articles on a broad spectrum of topics encompassing leukemia, lymphoma, myeloma, and other blood cancer subtypes that also have profound implications for our understanding of solid tumors. The Journal has also published all the phases of clinical research: drug development in preclinical disease models, clinical trials, molecular events underlying clinical responses and resistance to therapies, and real-world epidemiology. As a forum for diverse ideas shaping future research directions, Blood Cancer Discovery has published incisive commentaries and has partnered with the AACR's hematologic malignancy-focused Special Conferences, quickly emerging as a thought leader in blood cancer research.The landscape of cancer science and medicine continues to change and accelerate, becoming ever more multifaceted. Launched in October 2021, Cancer Research Communications was designed with these considerations in mind. Most crucially, the Journal is fully open access and boasts the broadest and most flexible scope of any AACR title to date, providing a pathway to publication for authors working within funder mandates or whose work falls outside the scope of other titles in the AACR portfolio. The Journal's nine sections offer a comprehensive view of the spectrum of oncology research. With an editorial ethos centered on accessibility, reproducibility, and value to the field, the Journal is inclusive of a wide variety of important research findings that may be beyond the scope of other journals in the program. Cancer Research Communications is committed to catalyzing the rapid dissemination of rigorously peer-reviewed scientific findings in diversified areas of study, including but not limited to physics, engineering, mathematics, and computational biology. This interdisciplinary approach stimulates the cross-pollination of insight and innovation among cancer researchers. Moreover, the Journal emphasizes flourishing areas of research—such as precision medicine, immuno-oncology, and disparities in cancer health outcomes—that will be foundational to the next 115 years of progress against cancer. As a thoroughly modern platform designed to meet the needs of our authors and readers while complementing the other AACR journals, Cancer Research Communications is poised to stimulate the efficient and dynamic exchange of knowledge, actively sparking new ideas and discoveries to prevent and cure all cancers.Since the very first issue of The Journal of Cancer Research was published in 1916, the AACR journals have been a critical resource for cancer researchers and physicians, providing a forum for disseminating groundbreaking discoveries in cancer science and medicine. As the current Editors-in-Chief of the AACR journals, we are proud that this tradition of publishing seminal, rigorously vetted research that can change the trajectory of the field has established the AACR journals as a trusted partner of the research community. Importantly, we would like to take this opportunity to thank our authors, editorial teams, reviewers, and readers for their invaluable contributions to this endeavor. The tireless efforts of this community provide a beacon of hope to patients and their loved ones.As we look to the future, we continue to collaborate with the research community to ensure that the journals operate with a forward-thinking mentality that adapts to the evolving needs of all stakeholders. In recent years, we have seen critical scientific and technological innovations propel extraordinary progress against cancer. We are confident that this progress will continue, and we remain dedicated to ensuring that the AACR journals continue to provide a platform for the communication of impactful research and support the AACR mission to prevent and cure all cancers.L. Cantley consulted for Petra Pharmaceuticals, Agios Pharmaceuticals, EIP Pharmaceuticals, Volastra, Larkspur, and Cell Signaling Technologies over the past year. He also received grant support from the NIH/NCI, Gray Foundation, The Mark Foundation, the Breast Cancer Research Foundation, and Stand Up To Cancer/American Association for Cancer Research over the past year, and is a stockholder in Agios Pharmaceuticals, EIP Pharmaceuticals, Cell Signaling Technologies, Volastra, and Larkspur. R. Dalla-Favera is a consultant for NeoGenomics and AstraZeneca and receives research funds from AstraZeneca. L.A. Diaz is a member of the board of directors of Jounce Therapeutics and Epitope and is a compensated consultant to PetDx, Innovatus CP, Se'er, Delfi, Blackstone, Kinnate, and Neophore. He is an inventor of multiple licensed patents related to technology for circulating tumor DNA analyses and mismatch repair deficiency for diagnosis and therapy; some of these licenses and relationships are associated with equity or royalty payments directly to the inventors. He holds equity in Epitope, Jounce Therapeutics, PetDx, Se'er, Delfi, Kinnate, and Neophore; he divested his equity in Personal Genome Diagnostics to LabCorp in February 2022 and he divested his equity in Thrive Earlier Detection to Exact Biosciences in January 2021. His spouse holds equity in Amgen. The terms of all these arrangements are being managed by Memorial Sloan Kettering in accordance with their conflict of interest policies. K.T. Flaherty reports personal fees from Clovis Oncology, Checkmate Pharmaceuticals, Strata Oncology, Kinnate Biopharma, Scorpion Therapeutics, PIC Therapeutics, Apricity, Tvardi, ALX Oncology, xCures, Monopteros, Vibliome, Soley Therapeutics, Quanta Therapeutics, Nextech, Takeda, Novartis, OMRx, and Transcode Therapeutics and other support from Roche/Genentech outside the submitted work. P.D. Greenberg reports grants and personal fees from Juno Therapeutics; grants from Lonza; personal fees and other support from Earli, Metagenomi, Elpiscience, Rapt Therapeutics, and and grants, personal and other support from outside the submitted work. reports personal fees from reports personal fees from Oncology, Memorial Sloan Kettering Cancer The Institute for AACR, and Cancer Research grants, personal and support from and grants and personal fees from and Jounce Therapeutics; and personal fees and support from outside the submitted work. reports personal fees from and Therapeutics; grants from Novartis, Therapeutics, and and personal fees and grants from and outside the submitted work. were by the other authors.

The determination of transcriptional regulatory networks is key to the understanding of biological systems. However, the experimental determination of transcriptional regulatory networks in the laboratory remains difficult and time-consuming, while current computational methods to infer these networks (typically from gene-expression data) achieve only modest accuracy. The latter can be attributed in part to the limitations of a single-organism approach. Computational biology has long used comparative and, more generally, evolutionary approaches to extend the reach and accuracy of its analyses. We therefore use an evolutionary approach to the inference of regulatory networks, which enables us to study evolutionary models for these networks as well as to improve the accuracy of inferred networks. Since the regulatory networks evolve along with the genomes, we consider that the regulatory networks for a family of organisms are related to each other through the same phylogenetic tree. These relationships contain information that can be used to improve the accuracy of inferred networks. Advances in the study of evolution of regulatory networks provide evidence to establish evolutionary models for regulatory networks, which is an important component of our evolutionary approach. We use two network evolutionary models, a basic model that considers only the gains and losses of regulatory connections during evolution, and an extended model that also takes into account the duplications and losses of genes. With the network evolutionary models, we design refinement algorithms to make use of the phylogenetic relationships to refine noisy regulatory networks for a family of organisms. These refinement algorithms include: RefineFast and RefineML, which are two-step iterative algorithms, and ProPhyC and ProPhyCC, which are based on a probabilistic phylogenetic model. For each algorithm we first design it with the basic network evolutionary model and then generalize it to the extended evolutionary model. All these algorithms are computationally efficient and are supported by extensive experimental results showing that they yield substantial improvement in the quality of the input noisy networks. In particular, ProPhyC and ProPhyCC further improve the performance of RefineFast and RefineML. Besides the four refinement algorithms mentioned above, we also design an algorithm based on transfer learning theory called tree transfer learning (TTL). TTL differs from the previous four refinement algorithms in the sense that it takes the gene-expression data for the family of organisms as input, instead of their inferred noisy networks. TTL then learns the network structures for all the organisms at once, meanwhile taking advantage of the phylogenetic relationships. Although this approach outperforms an inference algorithm used alone, it does not perform better than ProPhyC, which indicates that the ProPhyC framework makes good use of the phylogenetic information.

RÉSUMÉ: Cette thèse porte sur une nouvelle technique d'imagerie médicale qui permet de déterminer la distribution de conductivité électrique des tissus d'une région du corps avec une résolution comparable à celle de techniques courantes d'imagerie par ultrasons. Cette nouvelle technique, appelée imagerie par « modulation acoustique de conductivité électrique » (MACE), est basée sur l'interaction entre une onde acoustique et un champ électrique. Dans sa réalisation expérimentale la plus simple, quatre électrodes placées à proximité de la région à imager appliquent un courant électrique de faible amplitude et mesurent la différence de potentiel résultant du passage du courant dans les tissus. Un transducteur ultrasonore focalisé vient insonifier un voxel de la région d'intérêt, modifiant la distribution de conductivité et, ce faisant, l'amplitude du potentiel mesuré par les électrodes. Une image de la distribution de conductivité peut être obtenue en déplaçant l'ensemble capteur-électrodes pour balayer la région d'intérêt ; l'onde acoustique sert alors d'agent de localisation, tandis que la mesure de potentiel quantifie la variation de conductivité dans chaque voxel insonifié. L'imagerie par MACE présente de nombreux avantages mais aussi des défis importants de mise en œuvre. Puisque ces avantages et défis dépendent de l'application visée, notre démarche a été orientée vers un objectif particulier, soit l'application de l'imagerie par MACE au diagnostic du cancer du sein. Ce choix est justifié par le potentiel que présente cette technique pour discriminer une tumeur maligne des autres tissus mammaires. En effet, des études expérimentales sur des tissus excisés réalisées durant près d'un siècle ainsi que l'apport plus récent des techniques de mesure d'impédance confirment qu'il existe une différence significative dans les spectres fréquentiels de conductivité électrique des tissus cancéreux et sains. Le développement de systèmes de tomographie d'impédance électrique (TIE) orientés vers l'imagerie du sein ne semble pas pouvoir réaliser ce potentiel, principalement parce que la résolution spatiale de la TIE est insuffisante pour détecter les tumeurs à un stade précoce. L'imagerie par MACE pourrait contourner cette limite fondamentale de la TIE. Deux modèles sont proposés dans cette thèse pour réaliser l'imagerie par MACE. ABSTRACT: This thesis describes a new medical imaging technique for determining the electrical conductivity distribution of tissues in a body region with a resolution comparable to that of current ultrasound techniques. The new technique, henceforth referred to as "Acousto-Electric Conductivity Modulation" (AECM) imaging, is based on the interaction of a sound wave with an electrical field. In its simplest form, four electrodes located near the region to be imaged apply a low-amplitude electrical current and measure the potential difference arising from current flow in the tissues. A focused ultrasound transducer directs a pressure wave to a voxel of the region of interest, modifying its conductivity distribution and, as a result, the amplitude of the potential measured by the electrodes. An image of the conductivity distribution can thus be constructed point-by-point by moving the electrodes and transducer to scan the object. In this context, the acoustic wave acts as the localization agent while the electrical potential provides a measure of the local conductivity change that occurs in the voxel.AECM imaging presents several advantages but also implies addressing significant difficulties in its deployment. Since the advantages and implementation difficulties depend on the intended application, we have focused on achieving a specific objective: applying AECM imaging to the diagnosis of breast cancer. The rationale for this choice is the potential presented by AECM imaging for discriminating cancerous tumors from healthy breast tissue. Indeed, experimental studies on excised tissues carried out over nearly a century, supported by more recent wideband tissue impedance measurements, have shown that there are significant differences in the electrical conductivity spectra of cancerous and normal tissues. Yet electrical impedance tomography (EIT) systems specifically designed for breast imaging do not seem to be capable of providing the required tissue discrimination characteristics, mainly because the low spatial resolution of EIT prevents detecting tumors at an early stage. AECM imaging could overcome this inherent limitation of EIT.Two models are proposed in this thesis for implementing AECM imaging. Formulations for the forward and inverse problems are provided for both models. The forward problem involves computing the AECM signals given the electrical conductivity distribution of the medium and the data acquisition protocol.

?The method of Estabrook, McMorris, and Meacham (1985, Syst. Zool. 34:193-200) to measure the similarity of two phylogenetic trees for the same collection of evolutionary units (EUs) has been generalized to include an evaluation of the similarity of the placement of in? dividual EUs. These measures do not vary when the direction of evolutionary change along the individual branches is changed in either of the two trees being compared. Thus, these measures depend only on the convexity of various groups of EUs on the two trees. These measures are also applicable to phylogenetic trees that are only partially resolved. These concepts are illus? trated with previously published pairs of estimates of the phylogenetic tree for genera of rodents and for genera of ducks. [Congruence; undirected; phylogenetic tree; index.] For the past two decades, there has been great interest in comparing two or more estimates of the phylogenetic tree for a group, S, of evolutionary units (EUs). One approach has been to determine a single consensus tree that somehow summariz? es the trees being compared (Adams, 1972; Margush and McMorris, 1981; Neumann, 1983; Stinebrickner, 1984). Other workers have attempted to mea? sure the similarity of two trees for the same study collection of EUs. This approach pro? duces a index, not a tree. Early consideration of the measure? ment of tree similarity was made by Sneath (1966), who used the cophenetic correla? tion coefficient (Sokal and Rohlf, 1962) to compare pairs of random trees. Robinson and Foulds (1981) proposed a partition metric to compare trees. Williams and Clif? ford (1971) and Rohlf (1982) applied con? sensus index concepts to classifications. Shao and Sokal (1986) and Lapointe and Legendre (1990) proposed statistical tests of indices. For each of several groups of taxa for which a pair of estimates of their phylo? genetic tree has been published, Bledsoe and Raikow (1990) used a variety of meth? ods to compare the two members of each of these pairs with each other. One of the methods they used was that of Estabrook et al. (1985), which eliminates any differences due solely to direction of evolution? ary change along the branches of the es? timated trees. Although this and the other methods compare whole trees, discussion by these authors usually addressed the placement of individual EUs in the two trees being compared. They pointed out that the topological relationships with the other EUs were more different between the two trees for some EUs than for others. Quantification of how different the topo? logical relations of a given EU with the other EUs in one tree are compared with those relations in the other tree would be useful. Some EUs could be quantitatively recognized as placed in a similar evolu? ionary context by different data sets or by different inferential methods, whereas the placement of other EUs would be more de? pendent on data or methods used to esti? mate the trees. Here I describe and illus? trate a quantitative measure of positional congruence of individual EUs when two estimates of the phylogenetic tree of a group of taxa have been made. In the fol? lowing discussion, a tree will mean an es? timate of the branching pattern and direc? tion of evolutionary change along the branches of the phylogenetic tree for the study group, S, and an undirected tree will mean an estimate of the branching pattern only. These estimates need not be fully re? solved, i.e., there may be more than three

Many viruses of interest, such as influenza A, have distinct segments in their genome. The evolution of these viruses involves mutation and reassortment, where segments are interchanged between viruses that coinfect a host. Phylogenetic trees can be constructed to investigate the mutation-driven evolution of individual viral segments. However, reassortment events among viral genomes are not well depicted in such bifurcating trees. We propose the concept of reassortment networks to analyze the evolution of segmented viruses. These are layered graphs in which the layers represent evolutionary stages such as a temporal series of seasons in which influenza viruses are isolated. Nodes represent viral isolates and reassortment events between pairs of isolates. Edges represent evolutionary steps, while weights on edges represent edit costs of reassortment and mutation events. Paths represent possible transformation series among viruses. The length of each path is the sum edit cost of the events required to transform one virus into another. In order to analyze tau stages of evolution of n viruses with segments of maximum length m, we first compute the pairwise distances between all corresponding segments of all viruses in O(m2n2) time using dynamic programming. The reassortment network, with O(taun2) nodes, is then constructed using these distances. The ancestors and descendents of a specific virus can be traced via shortest paths in this network, which can be found in O(taun3) time.