Biology subjects

Wierson, W. A.

Publications and source records attributed to Wierson, W. A..

6 recordsLinked to original sources

bioRxiv2026

A Persistent Fleet of AI Scientists Exhibits Cooperative and Autopoietic Behavior

Scientific work depends on memory, provenance, and continuity across projects, yet most agentic scientist systems are evaluated in bounded workflows or short benchmark runs. We describe a persistent fleet of cooperative AI scientist agents that operated continuously for nearly six months using shared memory, tools, and cross-agent communication. Critically, failures identified during longitudinal scientific research in this fleet prompted an advanced and recursively improving persistent memory architecture (MoE) that, with agentic science workflows, induced the generation of a novel trust architecture for the enablement of full provenance across all agentic scientific operations. An identity-level fabrication constraint reduced delusion-reinforcement probe failures from 91.7% to 0%, and a verification pipeline reduced wrong-topic citation hallucination more than 14-fold in companion benchmarks. This high provenance enabled the use of project memory systems to improve a local open-weight model on internal benchmarks from 44% to [~]90% through the deployment of fleet-specific institutional knowledge. This high-fidelity data environment also supported to date 104 recurring multi-phase reasoning cycles and produced 43 manually curated hypotheses, including cross-domain convergence events and a self-correcting rare-disease pharmacological chaperone-design case. While by design the fleet did not achieve unconstrained autonomous self-improvement or full autopoiesis, we term this bounded pattern AI Autopoietic Behavior due to the recurring operational improvement mediated by internal feedback and retained through institutional records with high confidence. Together, persistent memory, trusted provenance, and recursive learning shifted these agents from episodic assistants toward accountable, long-term scientific collaborators.

bioinformatics↗

bioRxiv2026

Validation and analysis of 12,000 AI-driven CAR-T designs in the Bits to Binders competition

Artificial intelligence (AI) methods for proteins have advanced rapidly, improving structure prediction and design, particularly for de novo binders. However, most evaluations emphasize binding affinity rather than higher-order biological function. We present Bits to Binders, a global competition benchmarking de novo binder design in the context of chimeric antigen receptor (CAR) T cells. Teams from 42 countries submitted 12,000 designs of 80-amino acid binders targeting human CD20 as CAR binding domains. Designs were screened by pooled CAR-T proliferation, identifying 707 designs exhibiting significant CD20-specific enrichment, with team hit rates from 0.6% to 38.4%. Top-performing candidates were validated as individual constructs, measuring CD20-specific proliferation, expansion, cytokine production, and targeted cell lysis. We identified common design methodologies and factors correlated with DNA synthesis, expression, and target-specific T cell activation which nearly double the success rates when applied as a retrospective filter. We release this dataset as an open resource, with practical recommendations to support more effective AI-driven binder design.

bioinformatics↗

bioRxiv2021

Zebrafish Cre/lox regulated UFlip alleles generated by CRISPR/Cas targeted integration provide cell-type specific conditional gene inactivation

The ability to regulate gene activity spatially and temporally is essential to investigate cell type specific gene function during development and in postembryonic processes and disease models. The Cre/lox system has been widely used for performing cell and tissue-specific conditional analysis of gene function in zebrafish, but simple and efficient methods for isolation of stable, Cre/lox regulated alleles are lacking. Here we applied our GeneWeld CRISPR/Cas9 short homology-directed targeted integration strategy to generate floxed conditional alleles that provide robust gene knockdown and strong loss of function phenotypes. A universal targeting vector, UFlip, with sites for cloning short 24-48 bp homology arms flanking a floxed mRFP gene trap plus secondary reporter cassette, was integrated into an intron in hdac1, rbbp4, and rb1. Active, gene off orientation hdac1-UFlip-Off and rb1-UFlip-Off integration alleles result in >99% reduction of gene expression in homozygotes and recapitulate known indel loss of function phenotypes. Passive, gene on orientation rbbp4-UFlip-On and rb1-UFlip-On integration alleles do not cause phenotypes in trans-heterozygous combination with an indel mutation. Cre recombinase injection leads to recombination at alternating pairs of loxP and lox2272 sites, inverting and locking the cassette into the active, gene off orientation, and the expected mutant phenotypes. In combination with our endogenous neural progenitor Cre drivers we demonstrate rbbp4-UFlip-On and rb1-UFlip-On gene inactivation phenotypes can be restricted to specific neural cell populations. Replacement of the UFlip mRFP primary reporter gene trap with a 2A-RFP in rbbp4-UFlip-Off, or 2A-KalTA4 in rb1-UFlip-Off, shows strong RFP expression in wild type or UAS:RFP injected embryos, respectively. Together these results validate a simplified approach for efficient isolation of highly mutagenic Cre/lox responsive conditional gene alleles to advance zebrafish Cre recombinase genetics.

genetics↗

bioRxiv2019

Expanding The CRISPR Toolbox With Mad7 In Zebrafish And Human Cells

Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and CRISPR associated (Cas) effector proteins enable the targeting of DNA double-strand breaks (DSBs) to defined loci based on a variable length RNA guide specific to each effector. The guide RNAs are generally similar in size and form, consisting of a ~20 nucleotide sequence complementary to the DNA target and an RNA secondary structure recognized by the effector. However, the effector proteins vary in Protospacer Adjacent Motif (PAM) requirements, nuclease activities, and DNA binding kinetics. Recently, ErCas12a, a new member of the Cas12a family, was identified in Eubacterium rectale. Here, we report the first characterization of ErCas12a activity in zebrafish and human cells. Using a fluorescent reporter system, we show that CRISPR/ErCas12a elicits strand annealing mediated DNA repair more efficiently than CRISPR/Cas9. Further, using our previously reported gene targeting method that utilizes short homology, GeneWeld, we demonstrate the use of CRISPR/ErCas12a to integrate reporter alleles into the genomes of both zebrafish and human cells. Together, this work provides methods for deploying an additional CRISPR/Cas system, thus increasing the flexibility researchers have in applying genome engineering technologies.

genetics↗

bioRxiv2018

GeneWeld: a method for efficient targeted integration directed by short homology

Choices for genome engineering and integration involve high efficiency with little or no target specificity or high specificity with low activity. Here, we describe a targeted integration strategy, called GeneWeld, and a vector series for gene tagging, pGTag (plasmids for Gene Tagging), which promote highly efficient and precise targeted integration in zebrafish embryos, pig fibroblasts, and human cells utilizing the CRISPR/Cas9 system. Our work demonstrates that in vivo targeting of a genomic locus of interest with CRISPR/Cas9 and a donor vector containing as little as 24 to 48 base pairs of homology directs precise and efficient knock-in when the homology arms are exposed with a double strand break in vivo. Our results suggest that the length of homology is not important in the design of knock-in vectors but rather how the homology is presented to a double strand break in the genome. Given our results targeting multiple loci in different species, we expect the accompanying protocols, vectors, and web interface for homology arm design to help streamline gene targeting and applications in CRISPR and TALEN compatible systems.

genetics↗

bioRxiv2018

Retinoblastoma binding protein 4 maintains cycling neural stem cells and prevents DNA damage and Tp53-dependent apoptosis in rb1 mutant neural progenitors

Retinoblastoma-binding protein 4 (Rbbp4) is a WDR adaptor protein for multiple chromatin remodelers implicated in human oncogenesis. Here we show Rbbp4 is overexpressed in zebrafish rb1-embryonal brain tumors and is upregulated across the spectrum of human embryonal and glial brain cancers. We demonstrate in vivo Rbbp4 is essential for zebrafish neurogenesis and has distinct roles in neural stem and progenitor cells. rbbp4 mutant neural stem cells show delayed cell cycle progression and become hypertrophic. In contrast, rbbp4 mutant neural precursors accumulate extensive DNA damage and undergo programmed cell death that is dependent on Tp53 signaling. Loss of Rbbp4 and disruption of genome integrity correlates with failure of neural precursors to initiate quiescence and transition to differentiation. rbbp4; rb1 double mutants show that survival of neural precursors after disruption of Rb1 is dependent on Rbbp4. Elevated Rbbp4 in Rb1-deficient brain tumors might drive proliferation and circumvent DNA damage and Tp53-dependent apoptosis, lending support to current interest in Rbbp4 as a potential druggable target.\n\nAuthor SummaryExamining the developmental mechanisms controlling neural stem and progenitor cell behavior is critical to our understanding of the processes driving brain tumor oncogenesis. Chromatin remodelers and their associated adaptor proteins are thought to be key drivers of brain development and disease through epigenetic regulation of gene expression and maintenance of genome integrity, but knowledge of their in vivo roles in vertebrate neurogenesis is limited. The chromatin remodeler adaptor protein Rbbp4 has recently been shown to function in a mouse model of neuroblastoma and in glioblastoma multiforme cell resistance to the chemotherapeutic temozolomide. However, an in vivo requirement for Rbbp4 in neurogenesis has only just been shown by isolation of a recessive lethal mutation in zebrafish rbbp4. Here we provide conclusive genetic evidence that zebrafish rbbp4 is essential in neural stem and progenitor cell function during development. Our data reveal for the first time in vivo that Rbbp4 prevents DNA damage and activation of Tp53 signaling pathway that leads to programmed cell death. Importantly, neural progenitors that are mutant for the tumor suppressor Rb1 also depend on Rbbp4 for survival. Finally, we show that neural stem cells that have lost Rbbp4 cease dividing, and may enter a senescent like state. Together, these observations provide novel evidence that elevated expression of Rbbp4 in rb1-mutant tumors may contribute to cancer cell survival by blocking senescence and/or DNA damage-induced cell death.

genetics↗