Biology subjects

Carlsson, L.

Publications and source records attributed to Carlsson, L..

5 recordsLinked to original sources

bioRxiv2026

Dyrk1a gene dosage controls bipolar cell development and retinal connectivity.

Mammalian retina development is governed by a core set of eye-field transcription factors that orchestrate tissue lamination and neuronal connectivity through tightly regulated genetic hierarchies. The LIM homeodomain gene Lhx2 functions as a master eye-field regulator and this study identifies Dyrk1a as a previously unrecognised effector of Lhx2-dependent programs during mouse retinal development. Conditional deletion of Dyrk1a in retinal progenitor cells led to increased apoptosis within the inner retina demonstrating a requirement for Dyrk1a in cell survival and lineage maintenance. Adult heterozygous mice exhibited dorsoventral reductions in bipolar cell number and indicated a gene-dosage-sensitive control of interneuron survival. Bipolar cell loss disrupted their mosaic organisation and resulted in diminished activity across both scotopic and photopic retinal pathways. Dyrk1a haploinsufficient mice also displayed pronounced disruption of inner plexiform layer stratification consistent with defective laminar targeting and synaptic partner integration. This study defines a gene-dosage-dependent mechanism through which transcriptional hierarchies regulate neuronal connectivity and function during mammalian retinal development.

developmental biology↗

bioRxiv2023

CPSign - Conformal Prediction for Cheminformatics Modeling

Conformal prediction has seen many applications in pharmaceutical science, being able to calibrate outputs of machine learning models and producing valid prediction intervals. We here present the open source software CPSign that is a complete implementation of conformal prediction for cheminformatics modeling. CPSign implements inductive and transductive conformal prediction for classification and regression, and probabilistic prediction with the Venn-ABERS methodology. The main chemical representation is signatures but other types of descriptors are also supported. The main modeling methodology is support vector machines (SVMs), but additional modeling methods are supported via an extension mechanism, e.g. DeepLearning4j models. We also describe features for visualizing results from conformal models including calibration and efficiency plots, as well as features to publish predictive models as REST services. We compare CPSign against other common cheminformatics modeling approaches including random forest, and a directed message-passing neural network. The results show that CPSign produces robust predictive performance with comparative predictive efficiency, with superior runtime and lower hardware requirements compared to neural network based models. CPSign has been used in several studies and is in production-use in multiple organizations. The ability to work directly with chemical input files, perform descriptor calculation and modeling with SVM in the conformal prediction framework, with a single software package having a low footprint and fast execution time makes CPSign a convenient and yet flexible package for training, deploying, and predicting on chemical data.

bioinformatics↗

bioRxiv2020

A cutting-edge approach unravels a novel role for CDK6 in leukemic progenitor cells

Studies of molecular mechanisms of hematopoiesis and leukemogenesis are hampered by the unavailability of progenitor cell lines that accurately mimic the situation in vivo. We now report a robust method to generate and maintain LSK (lin-, Sca-1+, c-Kit+) cells which closely resemble MPP1 cells. HPCLSK reconstitute hematopoiesis in lethally irradiated recipient mice over more than eight months. Upon transformation with different oncogenes including BCR/ABL, FLT3-ITD or MLL-AF9 their leukemic counterparts maintain stem cell properties in vitro and recapitulate leukemia formation in vivo. The method to generate HPCLSK can be applied to transgenic mice and we illustrate it for CDK6-deficient animals. Upon BCR/ABLp210 transformation, Cdk6-/- HPCLSKs induce disease with a significantly enhanced latency and reduced incidence, showing the importance of CDK6 in leukemia formation. Studies of the CDK6 transcriptome in murine HPCLSK and human BCR/ABL+ cells have verified that certain pathways depend on CDK6 and have uncovered a novel CDK6-dependent signature, suggesting a role for CDK6 in leukemic progenitor cell homing. Loss of CDK6 may thus lead to a defect in homing. The HPCLSK system represents a unique tool for combined in vitro and in vivo studies and enables the production of large quantities of genetically modifiable hematopoietic or leukemic stem/progenitor cells. Key pointsO_LIWe describe the generation of murine cell lines (HPCLSK) which reliably mimic hematopoietic/leukemic progenitor cells. C_LIO_LICdk6-/- BCR/ABLp210 HPCLSKs uncover a novel role for CDK6 in homing. C_LI

cancer biology↗

bioRxiv2020

Local retinoic acid directs emergence of the extraocular muscle functional unit

Coordinated development of muscles, tendons, and their attachment sites ensures emergence of functional musculoskeletal units that are adapted to diverse anatomical demands among different species. How these different tissues are patterned and functionally assembled during embryogenesis is poorly understood. Here, we investigated the morphogenesis of extraocular muscles (EOMs), an evolutionary conserved cranial muscle group that is crucial for the coordinated movement of the eyeballs and for visual acuity. By means of lineage analysis, we redefined the cellular origins of periocular connective tissues interacting with the EOMs, which do not arise exclusively from neural crest mesenchyme as previously thought. Using 3D imaging approaches, we established an integrative blueprint for the EOM functional unit. By doing so, we identified a developmental time window where individual EOMs emerge from a unique muscle anlage and establish insertions in the sclera, which sets these muscles apart from classical muscle-to-bone type of insertions. Further, we demonstrate that the eyeballs are a source of diffusible retinoic acid that allow their targeting by the EOMs in a temporal and dose dependent manner. Using genetically modified mice and inhibitor treatments, we find that endogenous local variations in the concentration of retinoids contribute to the establishment of tendon condensations and attachment sites that precede the initiation of muscle patterning. Collectively, our results highlight how global and site-specific programs are deployed for the assembly of muscle functional units with precise definition of muscle shapes and topographical wiring of their tendon attachments.

developmental biology↗

bioRxiv2017

Applying Mondrian Cross-Conformal Prediction to Estimate Prediction Confidence on Large Imbalanced Bioactivity Datasets

Conformal prediction has been proposed as a more rigorous way to define prediction confidence compared to other application domain concepts that have earlier been used for QSAR modelling. One main advantage of such a method is that it provides a prediction region potentially with multiple predicted labels, which contrasts to the single valued (regression) or single label (classification) output predictions by standard QSAR modelling algorithms. Standard conformal prediction might not be suitable for imbalanced datasets. Therefore, Mondrian cross-conformal prediction (MCCP) which combines the Mondrian inductive conformal prediction with cross-fold calibration sets has been introduced. In this study, the MCCP method was applied to 18 publicly available datasets that have various imbalance levels varying from 1:10 to 1:1000 (ratio of active/inactive compounds). Our results show that MCCP in general performed well on cheminformatics datasets with various imbalance levels. More importantly, the method not only provides confidence of prediction and prediction regions compared to standard machine learning methods, but also produces valid predictions for the minority class. In addition, a compound similarity based nonconformity measure was investigated. Our results demonstrate that although it gives valid predictions, its efficiency is much worse than nonconformity measures obtained from supervised learning.

pharmacology and toxicology↗