Biology subjects

Englund, M.

Publications and source records attributed to Englund, M..

10 recordsLinked to original sources

bioRxiv2026

DecoderTCR: Compositional Pretraining and Entropy-Guided Decoding for TCR-pMHC Interactions

Modeling recognition between T-cell receptors (TCRs) and peptide-MHC (pMHC) complexes is a fundamental challenge in computational immunology, constrained by sparse paired interaction data relative to abundant unpaired sequences. We introduce DecoderTCR, a masked language model framework that addresses this through two contributions: (1) a compositional continual pre-training curriculum that learns component representations from marginal data before refining cross-chain dependencies from limited pairs, and (2) Iterative Entropy-Guided Refinement (IEGR), a non-autoregressive decoding algorithm that resolves high-confidence positions first to provide context for uncertain regions. On held-out benchmarks, DecoderTCR achieves 0.96 AUROC for zero-shot pMHC binding prediction and 0.76 AUROC for epitope-specific TCR recognition, approaching supervised baselines without epitope-specific training. Learned representations recover structural contacts without coordinate supervision, and generated sequences exhibit realistic recombination statistics. Experimental validation reveals a prediction-generation gap: strong discrimination does not yet yield reliable generation, highlighting an open challenge for the field.

bioinformatics↗

bioRxiv2025

Characterization of calcifications in posterior horn of human meniscus using micro-computed tomography

ObjectiveMeniscal calcifications are associated with meniscal degeneration and osteoarthritis (OA). We investigated micro-computed tomography ({micro}CT) imaging for identification of calcification patterns caused by basic calcium phosphate (BCP) and calcium pyrophosphate (CPP). DesignPosterior horns of human medial and lateral menisci from 19 individuals with medial compartment knee OA and 21 deceased donors were imaged with high-resolution {micro}CT. Raman spectroscopy characterized the calcification types from histological sections, adjacent to {micro}CT piece. Qualitative and quantitative analysis, visualization, and grading were performed using 3D {micro}CT images of meniscal calcifications. ResultsDifferent calcification patterns were observed with BCP and CPP. BCP was found at the borders of meniscal tissue and inside complex tears or fibrillation. In contrast, CPP accumulated as circumferential rod-like structures between collagen bundles, mainly inside the meniscal tissue, and lacked the porous 3D structure observed in BCP calcifications. Quantitatively, CPP samples had a higher calcification volume compared to BCP with a geometric mean ratio of 14 (95%CI; 3,73), and larger particle sizes with a ratio of 7 (95%CI; 2,8). Additionally, BCP calcifications had an organized porous structure with a closed porosity range of 6-19, while a similar structure was not seen in CPPs (range 1-3.5). ConclusionsWe qualitatively and quantitatively identified volumetric and morphological differences in the calcification deposition patterns between BCP and CPP calcifications in human meniscus. The calculated differences may help distinguish the calcification types with in vivo imaging modalities in the future, as well as provide a better understanding of their role in OA.

bioengineering↗

bioRxiv2025

Complex three-dimensional rearing environments amplify compensatory plasticity following early blindness

The neocortex has a remarkable capacity to alter its functional organization and connectivity in response to sensory loss, particularly if this loss occurs early in life. A key question is whether this cross-modal reorganization is driven by sensory deprivation or by enhanced use of the spared senses. We investigated how different rearing environments shape neural responses in primary somatosensory cortex (S1) of short-tailed opossums (Monodelphis domestica), following elimination of visual inputs through bilateral enucleation in early development. Early blind and sighted littermates were reared in enriched environments to promote active tactile exploration in three-dimensional (3D) space, or in standard laboratory cages. In adulthood, both enriched groups showed adaptive changes in exploration patterns and gap crossing behaviors relative to standard-reared counterparts. Thus, early blind animals showed behavioral compensation when challenged by complex environments. Enriched rearing increased selectivity of S1 neural responses to whisker touch and altered receptive field shapes such that they were less horizontally anisotropic. This shift was strongest in enriched early blind animals, enhancing tuning along the behaviorally relevant horizontal axis more than in standard-reared early blind animals. Thus, alterations in receptive fields of neurons in S1 following early blindness were amplified by environmental complexity. Sighted opossums reared with enrichment also showed similar whisker receptive field plasticity, though to a slightly lower degree. Together, these results demonstrate the strong influence of the rearing environment on reorganization of cortex that processes inputs from the spared senses, underscoring the role of experience in directing compensatory plasticity following early sensory loss. SIGNIFICANCE STATEMENTEnhanced perceptual abilities following early sensory loss are often attributed to cross-modal recruitment of cortex linked to the deprived sense. However, plasticity also occurs within cortical areas representing spared modalities. It remains unresolved whether deprivation alone is sufficient to induce such reorganization, or whether experience using the spared sense is required. We show that enriched rearing amplifies neural coding changes in primary somatosensory cortex after early blindness, shaping receptive field geometry and promoting adaptive behavioral strategies aligned with environmental demands. Comparable changes in sighted animals reared under the same conditions reveal that reliance on touch--rather than visual deprivation alone--drives this neural and behavioral plasticity, supporting the critical role of experience in enhancing functional outcomes after early sensory impairment.

neuroscience↗

bioRxiv2025

Characterizing meniscal calcifications with photon counting-based dual-energy computed tomography

Meniscal calcifications are associated with meniscal degeneration and osteoarthritis (OA). However, differentiating between calcification types, such as basic calcium phosphate (BCP) and calcium pyrophosphate (CPP), remains challenging in vivo. Therefore, assessing new imaging possibilities is crucial in understanding the calcification pathologies and their relationship with OA. This study investigated dual-energy computed tomography combined with a photon counting detector (PCD-DECT) to identify calcifications in human meniscal samples ex vivo, using Raman spectroscopy as reference. 82 meniscus samples from 41 donors were imaged using PCD-DECT at 120kVp. Data were collected in two energy bins (20-50keV and 50-120keV) with an isotropic resolution of 37{micro}m. Raman spectroscopy was used to classify the calcifications as BCP or CPP. Among 82 samples, Raman spectroscopy identified 36 samples with only either BCP or CPP calcifications, that were included in the subsequent analysis. Regression models were used to compare the dual-energy index (DEI) and to assess the low energy values for corresponding high energy values between calcification types. The highest difference observed between BCP and CPP in comparison of low energy values for corresponding high energy values was 166HU (95%CI: 73, 259) at high energy value of 500HU and the difference between DEI values was 0.035 (95%CI: 0.011, 0.059), suggesting a potential difference in the measured parameters of calcification types. To conclude, PCD-DECT allowed the measurement of BCP and CPP calcifications ex vivo, offering new potential in vivo applications in the future, which could help understand calcification processes and evaluate the efficacy of disease-modifying drugs targeting calcification inhibition.

bioengineering↗

bioRxiv2024

Species-associated bacterial diversity increases along a gradient of habitat degradation

Alterations of microbial communities have evident impacts on development, digestion, fecundity, metabolism, immunity, and diverse other biological functions of their hosts. Yet, the factors affecting microbial communities associated with wild species often remain uncharacterized. For example, the impact of the hosts habitat degradation due to anthropogenic activities has received little attention, which contrasts with the large literature showing how such habitat degradation is at least partly responsible for the on-going global patterns of macro-biodiversity erosion. Here, we use metacommunities of herbivorous insect species specialized in feeding on Plantago lanceolata in the fragmented landscape of the [A]land Islands, Finland, as a model system to test whether and how bacterial communities associated with wild species change along a gradient of habitat degradation. We evaluated microbial species diversity and community composition in two sympatric insect species sampled from local meadow habitats with various levels of human disturbance within or around these habitats (e.g. forests, roads, agriculture fields, buildings). Counter to our expectations, we found that bacterial diversity can increase with habitat degradation, with individuals from more degraded habitats hosting more rare bacterial species. In contrast, as the dominant microbial species remain similar across habitats, the community composition and function of the microbiota persist under habitat degradation. In this system, the strength of human activities might induce changes in habitat heterogeneity rather than changes in overall habitat quality, thus allowing local insects to encounter and host more rare microbes rather than trigger local microbial extinction.

microbiology↗

bioRxiv2023

Altered co-expression patterns of synovial fluid proteins related to the immune system and extracellular matrix organization in late stage OA, compared to non-OA controls

ObjectiveSynovial fluid contains proteins that may have been released from surrounding tissues, our aim was to gain new insights into the proteomic profiles of human synovial fluid in knees with and without osteoarthritis (OA). MethodsWe used synovial fluid from 11 patients with end-stage medial compartment knee OA, aspirated during total knee replacement, and from 13 deceased donors who had no prior history of knee OA (healthy controls). These samples were analyzed using high-multiplex immunoassays Olink(R). The differential expression of proteins between the groups was analyzed using a linear mixed effects model. The linear associations between pairs of protein expressions were estimated with a linear regression model. ResultsWe found that almost half of the detected proteins were differentially expressed between the OA and non-OA controls. The proteins that were most elevated in the OA group compared to controls were tartrate-resistant acid phosphatase type 5 (fold change 10.6, 95% CI [6.6-17.0]), coagulation factor XI (4.3 [2.6-6.8]) and urokinase-type plasminogen activator (4.3 [2.3-6.8]). The proteins with lower levels in OA compared to controls were fatty acid-binding protein, adipocyte (0.03 [0.02-0.05]), myocilin (0.05 [0.03-0.08]) and carbonic anhydrase 3 (0.14 [0.09-0.23]). The protein-protein co-expression analysis suggests an overall lower number of protein pairs that show co-expression in OA. ConclusionThere is a substantial change in protein abundance in synovial fluid in end-stage knee OA, suggesting that global joint homeostasis is severely deranged. Our findings suggest altered co-expression between the immune response and extracellular matrix organization in end-stage knee OA, in comparison to non-OA controls.

molecular biology↗

bioRxiv2022

Associations of human femoral condyle cartilage structure and composition with viscoelastic and constituent-specific material properties at different stages of osteoarthritis

The relationships between structure and function in human knee femoral cartilage are not well-known at different stages of osteoarthritis. Thus, we characterized the depth-dependent composition and structure of normal and osteoarthritic human femoral condyle cartilage (n = 47) and related them to their viscoelastic and constituent-specific mechanical properties. We observed that, in superficial cartilage, the collagen network disorganization and proteoglycan loss were associated with the smaller initial fibril network modulus (collagen pretension). Furthermore, the proteoglycan loss was associated with the greater strain-dependent fibril network modulus (a measure of nonlinear mechanical behavior). The proteoglycan loss was also associated with greater cartilage viscosity at a low loading frequency (0.005 Hz), while the disorganization of the collagen network was associated with greater cartilage viscosity at a high loading frequency (1 Hz). Our results suggest that proteoglycan degradation and collagen disorganization reduce the pretension of the collagen network while proteoglycan degradation also increases the nonlinear mechanical response of the collagen network. Further, the results also highlight that proteoglycan degradation and collagen disorganization increase the viscosity of cartilage, but their contribution to increased viscosity occurs in completely different loading frequencies.

biophysics↗

bioRxiv2021

Mineral crystal thickness in calcified cartilage and subchondral bone in healthy and osteoarthritic knees

Osteoarthritis (OA) is the most common joint disease globally. In OA, articular cartilage degradation is often accompanied with sclerosis of the subchondral bone. However, the association between OA and tissue mineralization at the nanostructural level is currently not understood. Especially, it is technically challenging to identify calcified cartilage, where relevant but poorly understood pathological processes like tidemark multiplication and advancement occur. Here, we used state-of-the-art micro-focus small-angle X-ray scattering with high 5{micro}m spatial resolution to determine mineral crystal thickness in human subchondral bone and calcified cartilage. Specimens with a wide spectrum of OA severities were acquired from the medial and lateral compartments of medial compartment knee OA patients (n=15) and cadaver knees (n=10). For the first time, we identified a well-defined layer of calcified cartilage associated with pathological tidemark multiplication, containing 0.32nm thicker crystals compared to the rest of calcified cartilage. In addition, we found 0.2nm thicker mineral crystals in both tissues of the lateral compartment in OA compared with healthy knees, indicating a loading-related disease process since the lateral compartment is typically less loaded in medial compartment knee OA. Furthermore, the crystal thickness of the subchondral bone was lower with increasing histopathological OA severity. In summary, we report novel changes in mineral crystal thickness during OA. Our data suggest that unloading in the knee is associated with the growth of mineral crystals, which is especially evident in the calcified cartilage. In the subchondral bone, mineral crystals become thinner with increasing OA severity, which indicates new bone formation with sclerosis. One Sentence SummaryMineral crystal thickness increases with osteoarthritis in the lateral condyle that is typically unloaded.

biophysics↗

bioRxiv2020

ProteoMill: Efficient network-based functional analysis portal for proteomics data

Functional analysis has become a common approach to incorporate biological knowledge into the analysis of omics data, and to explore molecular events that govern a disease state. It is though only one step in a wider analytical pipeline that typically requires use of multiple individual analysis software. There is currently a need for a well-integrated omics analysis tool that performs all the steps. The ProteoMill portal is developed as an R Shiny application and integrates all necessary steps from data-upload, converting identifiers, to quality control, differential expression and network-based functional analysis into a single fast, interactive easy to use workflow. Further, it maintains annotation data sources up to date, overcoming a common problem with use of outdated information, and seamlessly integrates multiple R-packages for an improved user-experience. The functionality provided in this software can benefit researchers by facilitating the exploratory analysis of proteomics data. ProteoMill is available for free at https://proteomill.com.

bioinformatics↗

bioRxiv2020

Proteomics profiling of human synovial fluid suggests global increased protein interplay in early-osteoarthritis (OA) and lost in late-stage OA

The underlying molecular mechanisms in osteoarthritis (OA) development are largely unknown. This study explores the proteome and the pairwise interplay of proteins on a global level in synovial fluid from patients with late-stage knee OA (arthroplasty), early knee OA (arthroscopy due to degenerative meniscal tear) and from deceased controls without knee OA. Synovial fluid samples were analyzed using state-of-the-art mass spectrometry with data-independent acquisition. The differential expression of the proteins detected was clustered and evaluated with data mining strategies and a multilevel model. Group-specific slopes of associations were estimated between expressions of each pair of identified proteins to assess the co-expression (i.e. interplay) between the proteins in each group. More proteins were increased in early-OA vs controls than late-stage OA vs controls. For most of these proteins, the fold changes between late-stage OA vs controls and early stage OA vs controls were remarkably similar suggesting potential involvement in the OA process. Further, for the first time this study illustrated distinct patterns in protein co-expression suggesting that the global interplay between the protein machinery is increased in early-OA and lost in late-stage OA. Further efforts should probably focus on earlier stages of the disease than previously considered.

molecular biology↗