Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
12
result(s) for
"Bayer, Fritz"
Sort by:
Network-based clustering unveils interconnected landscapes of genomic and clinical features across myeloid malignancies
Myeloid malignancies exhibit considerable heterogeneity with overlapping clinical and genetic features among subtypes. We present a data-driven approach that integrates mutational features and clinical covariates at diagnosis within networks of their probabilistic relationships, enabling the discovery of patient subgroups. A key strength is its ability to include presumed causal directions in the edges linking clinical and mutational features, and account for them aptly in the clustering. In a cohort of 1323 patients, we identify subgroups that outperform established risk classifications in prognostic accuracy. Our approach generalises well to unseen cohorts with classification based on our subgroups similarly offering advantages in predicting prognosis. Our findings suggest that mutational patterns are often shared across myeloid malignancies, with distinct subtypes potentially representing evolutionary stages en route to leukemia. With pancancer TCGA data, we observe that our modelling framework extends naturally to other cancer types while still offering improvements in subgroup discovery.
Myeloid malignancies vary significantly in their clinical outcomes and their genetic background. Here, the authors develop a network-based clustering method to predict subgroups of malignancies across disease subtypes.
Journal Article
Unbiased Signal Equation for Quantitative Magnetization Transfer Mapping in Balanced Steady-State Free Precession MRI
Purpose: Quantitative magnetization transfer (qMT) imaging can be used to quantify the proportion of protons in a voxel attached to macromolecules. Here, we show that the original qMT balanced steady-state free precession (bSSFP) model is biased due to over-simplistic assumptions made in its derivation. Theory and Methods: We present an improved model for qMT bSSFP, which incorporates finite radio-frequency (RF) pulse effects as well as simultaneous exchange and relaxation. Further, a correction to finite RF pulse effects for sinc-shaped excitations is derived. The new model is compared to the original one in numerical simulations of the Bloch-McConnell equations and in previously acquired in-vivo data. Results: Our numerical simulations show that the original signal equation is significantly biased in typical brain tissue structures (by 7-20 %) whereas the new signal equation outperforms the original one with minimal bias (< 1%). It is further shown that the bias of the original model strongly affects the acquired qMT parameters in human brain structures, with differences in the clinically relevant parameter of pool-size-ratio of up to 31 %. Particularly high biases of the original signal equation are expected in an MS lesion within diseased brain tissue (due to a low T2/T1-ratio), demanding a more accurate model for clinical applications. Conclusion: The improved model for qMT bSSFP is recommended for accurate qMT parameter mapping in healthy and diseased brain tissue structures.
Paper
Fair Clustering: A Causal Perspective
Clustering algorithms may unintentionally propagate or intensify existing disparities, leading to unfair representations or biased decision-making. Current fair clustering methods rely on notions of fairness that do not capture any information on the underlying causal mechanisms. We show that optimising for non-causal fairness notions can paradoxically induce direct discriminatory effects from a causal standpoint. We present a clustering approach that incorporates causal fairness metrics to provide a more nuanced approach to fairness in unsupervised learning. Our approach enables the specification of the causal fairness metrics that should be minimised. We demonstrate the efficacy of our methodology using datasets known to harbour unfair biases.
Paper
High-Dimensional Inference in Bayesian Networks
Inference of the marginal probability distribution is defined as the calculation of the probability of a subset of the variables and is relevant for handling missing data and hidden variables. While inference of the marginal probability distribution is crucial for various problems in machine learning and statistics, its exact computation is generally not feasible for categorical variables in Bayesian networks due to the NP-hardness of this task. We develop a divide-and-conquer approach using the graphical properties of Bayesian networks to split the computation of the marginal probability distribution into sub-calculations of lower dimensionality, thus reducing the overall computational complexity. Exploiting this property, we present an efficient and scalable algorithm for calculating the marginal probability distribution for categorical variables. The novel method is compared against state-of-the-art approximate inference methods in a benchmarking study, where it displays superior performance. As an immediate application, we demonstrate how our method can be used to classify incomplete data against Bayesian networks and use this approach for identifying the cancer subtype of kidney cancer patient samples.
Paper
Network-based clustering unveils interconnected landscapes of genomic and clinical features across myeloid malignancies
Myeloid malignancies exhibit considerable heterogeneity with overlapping clinical and genetic features among different subtypes. Current classification schemes, predominantly based on clinical features, fall short of capturing the complex genomic landscapes of these malignancies. Here, we present a data-driven approach that integrates mutational features and clinical covariates within networks of their probabilistic relationships, enabling the discovery of de novo cancer subgroups. In a cohort of 1323 patients across acute myeloid leukemia, myelodysplastic syndromes, chronic myelomonocytic leukemia and myeloproliferative neoplasms, we identified novel subgroups that outperform established risk classifications in prognostic accuracy. Our findings suggest that mutational patterns are often shared across different types of myeloid malignancies, with distinct subtypes potentially representing evolutionary stages en route to leukemia. Within the novel subgroups, our integrative method discerns unique patterns combining genomic and clinical features to provide a comprehensive view of the multifaceted genomic and clinical landscape of myeloid malignancies. This in turn may guide the development of targeted therapeutic strategies and offers a pathway to enhanced patient stratification.
Paper
Within-patient genetic diversity of SARS-CoV-2
Abstract SARS-CoV-2, the virus responsible for the current COVID-19 pandemic, is evolving into different genetic variants by accumulating mutations as it spreads globally. In addition to this diversity of consensus genomes across patients, RNA viruses can also display genetic diversity within individual hosts, and co-existing viral variants may affect disease progression and the success of medical interventions. To systematically examine the intra-patient genetic diversity of SARS-CoV-2, we processed a large cohort of 3939 publicly-available deeply sequenced genomes with specialised bioinformatics software, along with 749 recently sequenced samples from Switzerland. We found that the distribution of diversity across patients and across genomic loci is very unbalanced with a minority of hosts and positions accounting for much of the diversity. For example, the D614G variant in the Spike gene, which is present in the consensus sequences of 67.4% of patients, is also highly diverse within hosts, with 29.7% of the public cohort being affected by this coexistence and exhibiting different variants. We also investigated the impact of several technical and epidemiological parameters on genetic heterogeneity and found that age, which is known to be correlated with poor disease outcomes, is a significant predictor of viral genetic diversity. Author Summary Since it arose in late 2019, the new coronavirus (SARS-CoV-2) behind the COVID-19 pandemic has mutated and evolved during its global spread. Individual patients may host different versions, or variants, of the virus, hallmarked by different mutations. We examine the diversity of genetic variants coexisting within patients across a cohort of 3939 publicly accessible samples and 749 recently sequenced samples from Switzerland. We find that a small number of patients carry most of the diversity, and that patients with more diversity tend to be older. We also find that most of the diversity is concentrated in certain regions and positions of the virus genome. In particular, we find that a variant reported to increase infectivity is among the most diverse positions. Our study provides a large-scale survey of within-patient diversity of the SARS-CoV-2 genome.
Paper
The roles of vision and proprioception in spatial tuning of sensory attenuation
When we touch ourselves, the pressure appears weaker compared to when someone else touches us, an effect known as sensory attenuation. Sensory attenuation is spatially tuned and does only occur if the positions of the touching and the touched body-party spatially coincide. Here, we ask about the contribution of visual or proprioceptive signals to determine self-touch. By using a 3D arm model in a virtual reality environment, we dissociated the visual from the proprioceptive arm signal. When a virtual arm was visible indicating self-touch, we found that sensory attenuation generalized across different locations. When no virtual arm was visible, we found sensory attenuation to be strongest when subjects pointed to the position where they felt their arm to be located. We conclude that the spatial tuning of tactile attenuation depends on which signal determines the occurrence of self-touch. When observers can see their hand, the visual signal dominates the proprioceptive determining self-touch in a single visual snapshot. When only the proprioceptive signal is available, the positions of the touching and the touched body-part must be separately estimated and subsequently compared if they overlap in anatomical space.
Journal Article
Effects of oligonucleotide adsorption on the physicochemical characteristics of a nanoparticle-based model delivery system for antisense drugs
Cationic polystyrene nanoparticles, as a model drug carrier system for nucleic acids, are capable of binding negatively charged oligonucleotides by multiple electrostatic interactions. The effect of the adsorption of phosphorothioate oligonucleotides on the physicochemical properties of the carrier system was investigated for uncoated and sterically stabilized latex particles. Turbidity measurements and photon-correlation spectroscopy indicate that the colloidal stability of the nanoparticle-oligonucleotide conjugates is influenced by the number of oligonucleotides adsorbed on the carrier. Especially in the case of the uncoated material, a destabilizing effect has been observed up to oligonucleotide concentrations of 2.7 μmol/g polymer. Strikingly, at higher concentrations the latexes exhibit colloidal stability similar to the oligonucleotide-free samples. These results were correlated to zeta-potential measurements demonstrating a reversal from positive to negative values of the zeta potential with increasing oligonucleotide concentration. The points of zero charge of the particles are in the region of maximum coagulation. These findings were compared to adsorption studies and calculations based on the random sequential adsorption model. It appears that at first the colloidal stability of the carrier systems is diminished with increasing oligonucleotide adsorption, while higher surface coverages lead to a significant reduction in coagulation. At the saturation level the surface coverage can be considered as a monolayer of \"side-on\" adsorbed molecules and the conjugates exhibit colloidal stability similar to the bare particles without adsorbed molecules.[PUBLICATION ABSTRACT]
Journal Article
