Immunocytochemical Labelling of Haematological Samples Using Monoclonal Antibodies

I reflect on my experience working with David Y. Mason in the Leukaemia Research Laboratories in the Nuffield Department of Pathology at the University of Oxford in the early 1980s. This was soon after the first monoclonal antibodies had been produced, which led to an exciting and productive time in biological discovery and pathology diagnostics. A specific focus in the laboratory was the development of immunoenzymatic staining methods that would enable monoclonal antibodies to be applied in diagnostic practice. This paper describes the work that led to the performance of immuno-alkaline phosphatase staining on blood and bone marrow smears, the success of which changed leukaemia diagnosis.

Mapping information-rich genotype-phenotype landscapes with genome-scale Perturb-seq

A central goal of genetics is to define the relationships between genotypes and phenotypes. High-content phenotypic screens such as Perturb-seq (pooled CRISPR-based screens with single-cell RNA-sequencing readouts) enable massively parallel functional genomic mapping but, to date, have been used at limited scales. Here, we perform genome-scale Perturb-seq targeting all expressed genes with CRISPR interference (CRISPRi) across >2.5 million human cells and present a framework to power biological discovery with the resulting genotype-phenotype map. We use transcriptional phenotypes to predict the function of poorly-characterized genes, uncovering new regulators of ribosome biogenesis (including CCDC86, ZNF236, and SPATA5L1), transcription (C7orf26), and mitochondrial respiration (TMEM242). In addition to assigning gene function, single-cell transcriptional phenotypes allow for in-depth dissection of complex cellular phenomena - from RNA processing to differentiation. We leverage this ability to systematically identify the genetic drivers and consequences of aneuploidy and to discover an unanticipated layer of stress-specific regulation of the mitochondrial genome. Our information-rich genotype-phenotype map reveals a multidimensional portrait of gene function and cellular behavior.

Multiscale profiling of enzyme activity in cancer

Diverse processes in cancer are mediated by enzymes, which most proximally exert their function through their activity. Methods to quantify enzyme activity, rather than just expression, are therefore critical to our ability to understand the pathological roles of enzymes in cancer and to harness this class of biomolecules as diagnostic and therapeutic targets. Here we present an integrated set of methods for measuring specific enzyme activities across the organism, tissue, and cellular levels, which we unify into a methodological hierarchy to facilitate biological discovery. We focus on proteases for method development and validate our approach through the study of tumor progression and treatment response in an autochthonous model of Alk-mutant lung cancer. To quantitatively measure activity dynamics over time, we engineered multiplexed, peptide-based nanosensors to query protease activity in vivo. Machine learning analysis of sensor measurements revealed dramatic protease dysregulation in lung cancer, including significantly enhanced proteolytic cleavage of one peptide, S1 (Padj < 0.0001), which returned to healthy levels within three days after initiation of targeted therapy. Next, to link these organism-level observations to the in situ context, we established a multiplexed assay for on-tissue localization of enzyme activity and pinpointed S1 cleavage to endothelial cells and pericytes of the tumor vasculature. Lastly, to directly link enzyme activity measurements to cellular phenotype, we designed a high-throughput method to isolate and characterize proteolytically active cells, uncovering profound upregulation of pro-angiogenic transcriptional programs in S1-positive cells. Together, these methods allowed us to discover that protease production by angiogenic vasculature responds rapidly to targeted therapy against oncogene-addicted tumor cells, identifying a highly dynamic interplay between tumor cells and their microenvironment. This work provides a generalizable framework to functionally characterize enzyme activity in cancer.

Scalable Species Tree Inference with External Constraints

Species tree inference under the multi-species coalescent (MSC) model is a basic step in biological discovery. Despite the developments in recent years of methods that are proven statistically consistent and that have high accuracy, large datasets create computational challenges. Although there is gener- ally some information available about the species trees that could be used to speed up the estimation, only one method, ASTRAL-J, a recent development in the ASTRAL family of methods, is able to use this information. Here we describe two new methods, NJst-J and FASTRAL-J, that can estimate the species tree given partial knowledge of the species tree in the form of a non-binary unrooted constraint tree.. We show that both NJst-J and FASTRAL-J are much faster than ASTRAL-J and we prove that all three methods are statistically consistent under the multi-species coalescent model subject to this constraint. Our extensive simulation study shows that both FASTRAL-J and NJst-J provide advantages over ASTRAL-J: both are faster (and NJst-J is particularly fast), and FASTRAL-J is generally at least as accurate as ASTRAL-J. An analysis of the Avian Phylogenomics project dataset with 48 species and 14,446 genes presents additional evidence of the value of FASTRAL-J over ASTRAL-J (and both over ASTRAL), with dramatic reductions in running time (20 hours for default ASTRAL, and minutes or seconds for ASTRAL-J and FASTRAL-J, respectively). Availability: FASTRAL-J and NJst-J are available in open source form at https://github.com/ RuneBlaze/FASTRAL-constrained and https://github.com/RuneBlaze/NJst-constrained. Locations of the datasets used in this study and detailed commands needed to reproduce the study are provided in the supplementary materials at http://tandy.cs.illinois.edu/baqiao-suppl.pdf.

In Search of a Unifying Concept in Human Diseases

Throughout the history of biological/medicine sciences, there has been opposing strategies to find solutions to complex human disease problems. Both empirical and deductive approaches have led to major insights and concepts that have led to practical preventive and therapeutic benefits for the human population. The classic definitions of “science” (to know) has been paired with the classic definition of technology (to do). One knew more as the technology developed, and that development was often based on science. In other words, one could do more if science could improve the technology. In turn, this made possible to know more science with improved technology. However, with the development of new technologies of today in biology and medicine, major advances have been made, such as the information from the Human Genome Project, genetic engineering techniques and the use of bioinformatic uses of sophisticated computer analyses. This has led to the renewed idea that Precision Medicine, while raising some serious ethical concerns, also raises the expectation of improved potential of risk predictions for prevention and treatment of various genetically and environmentally influenced human diseases. This new field Artificial Intelligence, as a major handmaiden to Precision Medicine, is significantly altering the fundamental means of biological discovery. However, can today’s fundamental premise of “Artificial Intelligence”, based on identifying DNA, as the primary nexus of human health and disease, provide the practical solutions to complex human diseases that involve the interaction of those genes with the broad spectrum of “environmental factors”? Will it be “precise” enough to provide practical solutions for prevention and treatments of diseases? In this “Commentary”, with the example of human carcinogenesis, it will be challenged that, without the integration of mechanistic and hypothesis-driven approaches with the “unbiased” empirical analyses of large numbers of data, the Artificial Intelligence approach with fall short.

Eyes Toward Tomorrow Program Enhancing Collaboration, Connections, and Community Using Bioinspired Design

The goal of our Eyes Toward Tomorrow Program is to enrich the future workforce with STEM by providing students with an early, inspirational, interdisciplinary experience fostering inclusive excellence. We attempt to open the eyes of students who never realized how much their voice is urgently needed by providing an opportunity for involvement, imagination, invention, and innovation. Students see how what they are learning, designing, and building matters to their own life, community, and society. Our program embodies convergence by obliterating artificially created, disciplinary boundaries to go far beyond STEM or even STEAM by including artists, designers, social scientists, and entrepreneurs collaborating in diverse teams using scientific discoveries to create inventions that could shape our future. Our program connects two recent revolutions by amplifying Bioinspired Design with the Maker Movement and its democratizing effects empowering anyone to innovate and change the world. Our course is founded in original discovery. We explain the process of biological discovery and the importance of scaling, constraints, and complexity in selecting systems for bioinspired design. By spotlighting scientific writing and publishing, students become more science literate, learn how to decompose a biology research paper, extract the principles, and then propose a novel design by analogy. Using careful, early scaffolding of individual design efforts, students build the confidence to interact in teams. Team building exercises increase self-efficacy and reveal the advantages of a diverse set of minds. Final team video and poster project designs are presented in a public showcase. Our program forms a student-centered creative action community comprised of a large-scale course, student-led classes, and a student-created university organization. The program structure facilitates a community of learners that shifts the students' role from passive knowledge recipients to active co-constructors of knowledge being responsible for their own learning, discovery, and inventions. Students build their own shared database of discoveries, classes, organizations, research openings, internships, and public service options. Students find next step opportunities so they can see future careers. Description of our program here provides the necessary context for our future publications on assessment that examine 21st century skills, persistence in STEM, and creativity.

The Specious Art of Single-Cell Genomics

Dimensionality reduction is standard practice for filtering noise and identifying relevant dimensions in large-scale data analyses. In biology, single-cell expression studies almost always begin with reduction to two or three dimensions to produce 'all-in-one' visuals of the data that are amenable to the human eye, and these are subsequently used for qualitative and quantitative analysis of cell relationships. However, there is little theoretical support for this practice. We examine the theoretical and practical implications of low-dimensional embedding of single-cell data, and find extensive distortions incurred on the global and local properties of biological patterns relative to the high-dimensional, ambient space. In lieu of this, we propose semi-supervised dimension reduction to higher dimension, and show that such targeted reduction guided by the metadata associated with single-cell experiments provides useful latent space representations for hypothesis-driven biological discovery.

The Molecular Subtyping Resource (MouSR): a user-friendly tool for rapid biological discovery from human or mouse transcriptional data

Generation of transcriptional data has dramatically increased in the last decade, driving the development of analytical algorithms that enable interrogation of the biology underpinning the profiled samples. However, these resources require users to have expertise in data wrangling and analytics, reducing opportunities for biological discovery by wet-lab users with a limited programming skillset. Although commercial solutions exist, costs for software access can be prohibitive for academic research groups. To address these challenges, we have developed an open source and user-friendly data analysis platform for on-the-fly bioinformatic interrogation of transcriptional data derived from human or mouse tissue, called MouSR. This internet-accessible analytical tool, https://mousr.qub.ac.uk/, enables users to easily interrogate their data using an intuitive point and click interface, which includes a suite of molecular characterisation options including QC, differential gene expression, gene set enrichment and microenvironmental cell population analyses from RNA-Seq. Users are provided with adjustable options for analysis parameters to generate results that can be saved as publication-quality images. To highlight its ability to perform high quality data analysis, we utilise the MouSR tool to interrogate our recently published tumour dataset, derived from genetically engineered mouse models and matched organoids, where we rapidly reproduced the key transcriptional findings. The MouSR online tool provides a unique freely-available option for users to perform rapid transcriptomic analyses and comprehensive interrogation of the signalling underpinning transcriptional datasets, which alleviates a major bottleneck for biological discovery.