Science Museum of Minnesota

Abstract Introduction: Women with dense breast tissue face a two-fold higher risk of breast cancer, yet mammography sensitivity is as low as 30% in this group, underscoring the need for more sensitive and accessible screening methods. Liquid-based biopsies for early breast cancer detection are emerging but currently remain out of reach for clinical use. Recent data on nucleotide assessment from plasma in breast cancer have been mixed with 87% sensitivity for late-stage disease (stage 3-4) but only 20% sensitivity for early-stage disease (stage 1-2). Proteomics is an exciting area for early cancer screening where improvements in sample preparation and equipment have enabled major advances in early cancer detection. This is particularly true for a deep proteome assay as the detection of proteins 8-9 orders of magnitude lower in abundance than common plasma proteins can be identified. Objective: This study evaluates the presence of breast cancer using label-free shotgun mass spectrometry-based proteomics on less than 1 ml of plasma in >1,100 women classified as either healthy or as newly diagnosed breast cancer patients with a focus on early-stage disease (stages 0-2). Methods: 363 banked blood samples from multiple sources were used to establish a training set consisting of 156 stage 0-4 treatment-naive breast cancer patients, 46 benign breast conditions, and 161 healthy controls. Samples were age and demographic matched between healthy, benign, and breast cancer cohorts. Samples were processed in a blinded manner using an automated and novel deep proteomic platform developed by Astrin Biosciences. This platform combines label-free quantitative mass spectrometry and a proprietary breast cancer-specific spectral library to extract features used to train a classifier to identify breast cancer patients. Blinded validation of this AI classifier was evaluated on the remaining 788 patient samples (healthy, n=374; benign, n=81; breast cancer, n=333) to confirm performance characteristics. Results: Preliminary analyses from our training set of deep proteomic profiling identified over 8,600 proteins in total and up to 7,500 proteins per patient (median 6,926 proteins) with a dynamic range spanning over 8 orders of magnitude. Differentially expressed proteins in early-stage breast cancer patients from healthy patient controls were also observed. Our AI classifier, using a leave-one-out cross-validation (LOOCV) approach on the interim cohort, separated healthy controls from breast cancer patients with an ROC of >95% and a specificity and sensitivity greater than 85% across all stages. Conclusion: We have developed a highly sensitive blood-based assay that utilizes deep proteomic profiling to identify distinctive cancer specific signatures in women who are undergoing screening for breast cancer. This work enables us to develop a protein-based classifier from plasma for early detection of breast cancer. Impact on screening strategies: This assay advances clinical diagnostics by identifying proteomic markers for early detection of breast cancer. This innovative approach enhances screening strategies for women with dense breasts who are at average or high-risk based on family history, specific genetic mutations such as BRCA1/2, race, or other factors. Acknowledgements: We thank all of the members at Astrin Biosciences who have contributed endless hours dedicated to ending cancer. We are also grateful to all the donors who generously contributed blood samples to this study. Citation Format: A. Horrmann, Y. Travadi, G. Schaap, K. Mallery, K. J. Kamalanathan, N. Bristow, C. Galeano-Garces, S. Y. Bae, A. Hesch, C. Rungkittikhun, B. R. Konety, J. M. Drake. Deep Proteomics and AI Classifier for Early Breast Cancer Detection [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PD5-03.

Despite the widespread use of mammography as the standard of care for breast cancer screening, its accuracy remains limited for select patient populations, such as women with high breast density. Liquid biopsy-based tests offer an accessible complement to conventional screening methods. Here, we conducted a case-control study to develop a plasma-based protein classifier to distinguish between those with early-stage breast cancer and healthy individuals. A total of 335 women, comprising 116 patients with newly diagnosed, treatment-naïve breast cancer (stage 0-2) and 219 healthy controls, had plasma samples collected and processed in a blinded manner using a sample preparation method coupled with semiquantitative, label-free mass spectrometry-based analysis. The median number of proteins detected per patient across breast cancer and healthy individuals was 6991 and 6818, respectively. A machine learning-based classifier was trained and validated on patient proteome profiles using a leave-one-out cross-validation approach to identify patients with breast cancer. The classifier achieved an area under the curve of 0.96 (95% CI, 0.93-0.97), with a sensitivity of 86.2% (95% CI, 78.8-91.3%) and a specificity of 90.4% (95% CI, 85.8-93.6%). In patients with breast cancer, the classifier retained >85% sensitivity regardless of breast density (low density: 87.2%, high density: 90.2%) at 90% specificity. Our workflow demonstrates the potential of plasma proteomics as a potent diagnostic tool in early-stage breast cancer screening.

Abstract We here explore the relationship between biological and cultural evolution in Cistothorus palustris (Marsh Wren), which across Canada and the United States can be easily differentiated into two geographic groups based on diagnostic western and eastern song types. The wren songs are culturally transmitted, as young males raised in the laboratory readily learn songs of either type. We find that, in cattail- and bulrush-dominated marshes across a > 400-km transition zone in southern Saskatchewan, males of both cultural traditions occur together and defend territories against each other; pure western and pure eastern singers are abundant, and relatively few males have learned songs of both phenotypes. To determine the genetic ancestry of singing males in these two cultural traditions, we generated reduced-representation genomic sequence data for birds from across this contact zone. We find exceptionally high genome-wide divergence between western and eastern singing Marsh Wrens from outside the contact zone (FST = 0.61), potentially the highest between any lineages that form a hybrid zone in the North American Great Plains. In the heart of the contact zone, genetic data indicate an overall low rate of hybrid ancestry, and ancestry is tightly coupled with male song type, suggesting that this cultural trait could play a role in maintaining a strong but imperfect degree of assortative mating. All 9 of the mixed singers (males with both western and eastern songs) that we sequenced were of predominantly eastern ancestry, with 6 of them having 100% eastern genotype. These mismatched individuals could play an outsized role in driving the ongoing generation of hybrid offspring and the introgression of eastern alleles into western populations. Taken together, these genomic and vocal datasets reveal that the western and eastern Marsh Wren song phenotypes represent deeply divergent parapatric lineages despite breeding side-by-side in the same marshes. Although all males are capable of learning the songs of the other genotype, they rarely do, and two cultural traditions representing two genetic ancestries appear to coexist stably in the Great Plains of southern Saskatchewan.

Diet‐induced obesity is associated with enhanced systemic inflammation that limits bone regeneration. HDAC inhibitors are currently being explored as anti‐inflammatory agents. Prior reports show that myeloid progenitor‐directed Hdac3 ablation enhances intramembranous bone healing in female mice. In this study, we determined if Hdac3 ablation increased intramembranous bone regeneration in mice fed a high‐fat/high‐sugar (HFD) diet. Micro‐CT analyses demonstrated that HFD‐feeding enhanced the formation of periosteal reaction tissue of control littermates, reflective of suboptimal bone healing. We confirmed enhanced bone volume within the defect of Hdac3‐ablated females and showed that Hdac3 ablation reduced the amount of periosteal reaction tissue following HFD feeding. Osteoblasts cultured in a conditioned medium derived from Hdac3‐ablated cells exhibited a four‐fold increase in mineralization and enhanced osteogenic gene expression. We found that Hdac3 ablation elevated the secretion of several chemokines, including CCL2. We then confirmed that Hdac3 deficiency increased the expression of Ccl2. Lastly, we show that the proportion of CCL2‐positve cells within bone defects was significantly higher in Hdac3‐deficient mice and was further enhanced by HFD. Overall, our studies demonstrate that Hdac3 deletion enhances intramembranous bone healing in a setting of diet‐induced obesity, possibly through increased production of CCL2 by macrophages within the defect.

Abstract The passerine superfamily Certhioidea lacks a complete phylogeny despite decades of recognition as a clade and extensive systematic work within all its constituent families. Here, we inferred a near-complete species-level phylogeny of Certhioidea from a molecular supermatrix, including the first comprehensive sampling of the wrens (Troglodytidae), and used this phylogeny to infer its biogeographic and diversification histories. We also inferred an expanded phylogeny including nearly 100 putative phylospecies previously documented in the literature, and we found that including this diversity had notable impacts on the inferred evolutionary history of Certhioidea. This phylospecies-level tree documented a few instances of species paraphyly, some previously described in the literature and some novel. We found that Certhioidea originated largely in Eurasia and dispersed into North America five times in the last 20 million years, including at the origin of the “New World certhioids,” wrens and gnatcatchers, a clade herein named Orthourae. After this initial dispersal event, both wrens and gnatcatchers diversified extensively across the hemisphere, with both lineages repeatedly crossing between continents. However, we detected no notable impact of the formation of the Isthmus of Panama on the frequency of dispersal events between North and South America. The inclusion of phylospecies altered this biogeographic inference in some portions of the tree but overall was largely consistent. With species-level sampling, we found that diversification rates within Certhioidea were largely constant through time with a detectable deceleration toward the present. By contrast, phylospecies-level sampling recovered a different diversification history with a significant rate increase at the crown node of Orthourae after dispersing into the Americas and increased speciation rates particularly within the genera Polioptila and Henicorhina. This largely resolved phylogeny for Certhioidea has yielded important insights into the evolutionary history of this group and provides a framework for future comparative work on this fascinating clade.

The majority of multi-cellular terrestrial life is found in tropical forests and is either an invertebrate or a plant: for decades ecologists have sought to understand why. As global change erodes the list of extant species on our planet quantifying what species remain, along with their origins and ecology, contributes to our ability to preserve ecosystem functioning and resilience. Here we study three feeding guilds of caterpillars (Lepidoptera) and seek to understand the drivers of their diet breadth across four diverse tropical plant genera in Papua New Guinea. Host specificity is central to biodiversity estimates and the resilience of ecological networks. Specifically, we calculate distance-based host specificity in relation to plant phylogenetic relationships alongside chemical and mechanical traits of leaves. In terms of chemical defenses, we focus on the major polyphenol groups, a compound class shared across many plant species. We refine our data exploration using food webs and ordinations to pick out specific traits of relevance to insect host specificity. Our results showed that the degree of specialization for caterpillars took the following order: phylogenetic>polyphenol>mechanical, such that insect specificity was explained best by host phylogeny and polyphenol chemistry in our study system. Leaf mining insects had higher host specificity than those feeding externally. Of the traits studied hexahydroxydiphenoyl derivatives, galloyl derivatives, trichome density, quinic acid derivatives, myricetins and successional index explained the most variation in overall insect community structure. Our findings build on earlier studies of New Guinean rainforest communities and add a mechanistic explanation to previous findings that host genera are functional islands for insect herbivores. Further, we demonstrate that different plant genera combine different defensive traits that appear to drive associated insect diversity. Our approach integrates trait data and phylogeny to explore dimensions of specialization and we welcome metabolomic studies that will provide more detailed explanations for insect-herbivore host use. Finally, chemical diversity is directly linked to organismal diversity and by studying a range of insect herbivore guilds we make a connection between feeding ecology and specialization that will help to predict species interactions and, potentially, the persistence of ecological networks.

Abstract The greenhouse gas methane (CH4) contributed to a warm climate that maintained liquid water and sustained Earth’s habitability in the Precambrian despite the faint young sun. The viability of methanogenesis (ME) in ferruginous environments, however, is debated, as iron reduction can potentially outcompete ME as a pathway of organic carbon remineralization (OCR). Here, we document that ME is a dominant OCR process in Brownie Lake, Minnesota (midwestern United States), which is a ferruginous (iron-rich, sulfate-poor) and meromictic (stratified with permanent anoxic bottom waters) system. We report ME accounting for ≥90% and >9% ± 7% of the anaerobic OCR in the water column and sediments, respectively, and an overall particulate organic carbon loading to CH4 conversion efficiency of ≥18% ± 7% in the anoxic zone of Brownie Lake. Our results, along with previous reports from ferruginous systems, suggest that even under low primary productivity in Precambrian oceans, the efficient conversion of organic carbon would have enabled marine CH4 to play a major role in early Earth’s biogeochemical evolution.