A5059242392- Gene expression and cancer classification
- Bioinformatics and Genomic Networks
- Single-cell and spatial transcriptomics
- Medical Imaging Techniques and Applications
- Diabetes and associated disorders
- Parkinson's Disease Mechanisms and Treatments
- RNA Research and Splicing
- Neurological disorders and treatments
- Cancer-related molecular mechanisms research
- Diabetes Management and Research
- Pancreatic function and diabetes
- Microbial Metabolic Engineering and Bioproduction
- Metabolomics and Mass Spectrometry Studies
- Immune cells in cancer
- Bariatric Surgery and Outcomes
- Biomedical Text Mining and Ontologies
- Viral Infections and Immunology Research
- Respiratory viral infections research
- Viral Infectious Diseases and Gene Expression in Insects
- Body Contouring and Surgery
- Obstructive Sleep Apnea Research
- Statistical Methods and Inference
- Viral gastroenteritis research and epidemiology
- SARS-CoV-2 and COVID-19 Research
- Neuroscience of respiration and sleep
University of Turku
2015-2025
Turku Centre for Computer Science
2020-2025
Åbo Akademi University
2017-2025
Turku PET Centre
2023-2025
Turku University Hospital
2023-2025
Turku Centre for Biotechnology
2015-2018
Differential expression analysis is one of the most common types analyses performed on various biological data (e.g. RNA-seq or mass spectrometry proteomics). It process that detects features, such as genes proteins, showing statistically significant differences between sample groups under comparison. A major challenge in choice an appropriate test statistic, different statistics have been shown to perform well datasets. To this end, reproducibility-optimized statistic (ROTS) adjusts a...
We compared five statistical methods to detect differentially expressed genes between two distinct single-cell populations. Currently, it remains unclear whether differential expression developed originally for conventional bulk RNA-seq data can also be applied analysis. Our results in three diverse comparison settings showed marked differences the different terms of number detections as well their sensitivity and specificity. They, however, did not reveal systematic benefits currently...
Abstract Consumer wearables and sensors are a rich source of data about patients’ daily disease symptom burden, particularly in the case movement disorders like Parkinson’s (PD). However, interpreting these complex into so-called digital biomarkers requires complicated analytical approaches, validating sufficient unbiased evaluation methods. Here we describe use crowdsourcing to specifically evaluate benchmark features derived from accelerometer gyroscope two different datasets predict...
Abstract We evaluate deconvolution methods, which infer levels of immune infiltration from bulk expression tumor samples, through a community-wide DREAM Challenge. assess six published and 22 community-contributed methods using in vitro silico transcriptional profiles admixed cancer healthy cells. Several predict most cell types well, though they either were not trained to all functional CD8+ T states or do so with low accuracy. address this gap, including deep learning-based approach, whose...
Type 1 diabetes is a complex heterogenous autoimmune disease without therapeutic interventions available to prevent or reverse the disease. This study aimed identify transcriptional changes associated with progression in patients recent-onset type diabetes.Whole-blood samples were collected as part of INNODIA at baseline and 12 months after diagnosis diabetes. We used linear mixed-effects modelling on RNA-seq data genes age, sex, progression. Cell-type proportions estimated from using...
Background. Recently, dynamic total-body positron emission tomography (PET) imaging has become possible due to new scanner devices. While clustering algorithms have been proposed for PET analysis already earlier, there is still little research systematically evaluating these processing of images. Materials and methods. Here, we compare the performance 15 unsupervised methods, including K-means either by itself or after principal component (PCA) independent (ICA), Gaussian mixture model...
Multiple methods have been proposed to estimate pathway activities from expression profiles, and yet, there is not enough information available about the performance of those methods. This makes selection a suitable tool for analysis difficult. Although based on simple gene lists remained most common approach, various that also consider structure emerged. To provide practical insight both list-based structure-based methods, we tested six different approaches in two case study settings...
Abstract Segmentation is a routine step in PET image analysis, and few automatic tools have been developed for it. However, excluding supervised methods with their own limitations, they are typically designed older, small images the implementations no longer publicly available. Here, we test if different commonly used building blocks of work large modern total-body images. Dynamic from five datasets evaluation purposes, tested algorithms cover wide range preprocessing approaches unsupervised...
Abstract The response to respiratory viruses varies substantially between individuals, and there are currently no known molecular predictors from the early stages of infection. Here we conduct a community-based analysis determine whether pre- or post-exposure factors could predict physiologic responses viral exposure. Using peripheral blood gene expression profiles collected healthy subjects prior exposure one four (H1N1, H3N2, Rhinovirus, RSV), as well up 24 h following exposure, find that...
Computational deconvolution is a time and cost-efficient approach to obtain cell type-specific information from bulk gene expression of heterogeneous tissues like blood. Deconvolution can aim either estimate type proportions or abundances in samples, how strongly each present expresses different genes, both tasks simultaneously. Among the two separate goals, estimation proportions/abundances widely studied, but less attention has been paid on defining profiles. Here, we address this gap by...
Enterovirus infections have been associated with the development of type 1 diabetes in multiple studies, but little is known about enterovirus-induced responses children at risk for developing diabetes. Our aim was to use genome-wide transcriptomics data characterise enterovirus-associated changes whole-blood samples from genetic susceptibility Longitudinal (356 total) collected 28 pairs increased were screened presence enterovirus RNA. Seven these detected as enterovirus-positive, each them...
Abstract Mobile health, the collection of data using wearables and sensors, is a rapidly growing field in health research with many applications. Deriving validated measures disease severity that can be used clinically or as outcome clinical trials, referred to digital biomarkers, has proven difficult. In part due complicated analytical approaches necessary develop these metrics. Here we describe use crowdsourcing specifically evaluate benchmark features derived from accelerometer gyroscope...
Pathway analysis is a common approach in diverse biomedical studies, yet the currently-available pathway tools do not typically support increasingly popular personalized analyses. Another weakness of methods their inability to handle challenging data with only modest group-based effects compared natural individual variation. In an effort address these issues, this study presents novel method PASI (Pathway Analysis for Sample-level Information) and demonstrates its performance on complex...
Clustering time activity curves of PET images have been used to separate clinically relevant areas the brain or tumours. However, image segmentation in multiorgan level is much less studied due available total-body data being limited animal studies. Now, new scanners providing opportunity acquire scans also from humans are becoming more common, which opens plenty interesting opportunities. Therefore, organ-level has important applications, yet it lacks sufficient research. In this proof...
A bstract Clustering time activity curves of PET images has been used to separate clinically relevant areas the brain or tumours. However, image segmentation in multi-organ level is much less studied due available total-body data being limited animal studies. Now new scanners providing opportunity acquire scans also from humans are becoming more common, which opens plenty interesting opportunities. Therefore, organ important applications, yet it lacks sufficient research. In this proof...
<title>Abstract</title> <bold>Backgrounds</bold> Obesity is associated with alterations in bone turnover markers (BTMs). However, the association between regional fat distribution and metabolism has received less attention. This study therefore aimed to identify which specific compartments (i.e., abdominal femoral subcutaneous fat, intra- extraperitoneal total visceral fat) exert most significant influence on circulating BTMs. <bold>Methods</bold> The comprised a cohort of individuals severe...
Although type 1 diabetes (T1D) is primarily a disease of the pancreatic beta-cells, understanding disease-associated alterations in whole pancreas could be important for improved treatment or prevention disease. We have characterized whole-pancreas gene expression patients with recently diagnosed T1D from Diabetes Virus Detection (DiViD) study and non-diabetic controls. Furthermore, another parallel dataset an additional laser-captured islets DiViD organ donors were analyzed together...
Summary We introduce a new method for Pathway Analysis of Longitudinal data (PAL), which is suitable complex study designs, such as longitudinal data. The main advantages PAL are the use pathway structures and suitability approach settings beyond currently available tools. demonstrate performance with three datasets related to early development type 1 diabetes, involving different designs only subtle biological signals. Transcriptomic proteomic represented among test An R package...