Biology subjects

Suomi, T.

Publications and source records attributed to Suomi, T..

7 recordsLinked to original sources

bioRxiv2026

ROTS 2.0: A reproducibility-driven framework for robust statistical modeling across diverse high-throughput omics study designs

Reproducibility is fundamental to reliable scientific discoveries. The reproducibility-optimized test statistic (ROTS) is a robust framework designed to identify reproducible features (e.g. genes or proteins) in high-dimensional differential expression analyses such as transcriptomics and proteomics. This is achieved by optimizing the reproducibility of feature rankings under resampling. While originally implemented for univariate settings, ROTS now accommodates multi-group comparisons, survival analysis, linear models, and linear mixed-effects models, broadening its applicability to more complex and clinically relevant experimental designs. Using diverse simulations, benchmark datasets, and real-world case studies, we demonstrate the benefits of ROTS reproducibility optimization compared to the corresponding conventional test statistics. Additionally, we illustrate the utility of the reproducibility characteristics in assessing the overall reliability of the results. To facilitate widespread adoption, ROTS is provided as an open-source software package available through R/Bioconductor. Furthermore, to broaden the user base, we now also provide a Python interface available at pypi.org/project/PyROTS/.

bioinformatics↗

bioRxiv2025

Limitations of de novo sequencing in resolving sequence ambiguity

De novo peptide sequencing enables peptide identification from fragmentation spectra without relying on sequence databases. However, incomplete spectra create ambiguity, making unambiguous identification challenging. Recent deep learning advances have produced numerous de novo models that predict sequences and refine peptide-spectrum matches under such conditions. Yet, their relative strengths, weaknesses, and ability to handle spectrum ambiguity remain unclear. Here, we benchmark eight state-of-the-art models on three publicly available proteomics datasets, comparing performance using established metrics and quantifying inter-model agreement. We assess post-processing approaches, including iterative refinement, rescoring, and reranking, for their ability to improve identification accuracy, and perform an error analysis to identify common mispredictions and their causes. Model performance varied, with considerable overlap of correct identifications. Post-processing yielded no or only modest improvements. Most sequencing errors were model-independent and driven by limited fragment ion coverage, a limitation also observed in database searches with large search spaces.

bioinformatics↗

bioRxiv2022

Longitudinal pathway analysis using structural information with case studies in early type 1 diabetes

We introduce a new method for Pathway Analysis of Longitudinal data (PAL), which is suitable for complex study designs, such as longitudinal data. The main advantages of PAL are the use of pathway structures and the suitability of the approach for study settings beyond currently available tools. We demonstrate the performance of PAL with three longitudinal datasets related to the early development of type 1 diabetes, involving different study designs and only subtle biological signals. Transcriptomic and proteomic data are represented among the test data. An R package implementing PAL is publicly available at https://github.com/elolab/PAL. MotivationPathway analysis is a frequent step in studies involving gene or protein expression data, but most of the available pathway methods are designed for simple case versus control studies of two sample groups without further complexity. The few available methods allowing the pathway analysis of more complex study designs cannot use pathway structures or handle the situation where the variable of interest is not defined for all samples. Such scenarios are common in longitudinal studies with so long follow up time that healthy controls are required to identify the effect of normal aging apart from the effect of disease development, which is not defined for controls. PAL is the first available pathway method to analyse such high-investment datasets.

bioinformatics↗

bioRxiv2021

PhosPiR: An automated phospho-proteomic pipeline in R

Large-scale phospho-proteome profiling using mass spectrometry (MS) provides functional insight that is crucial for disease biology and drug discovery. However, extracting biological understanding from this data is an arduous task requiring multiple analysis platforms that are not adapted for automated high-dimensional data analysis. Here, we introduce an integrated pipeline that combines several R packages to extract high-level biological understanding from largescale phosphoproteomic data by seamless integration with existing databases and knowledge resources. In a single run, PhosPiR provides data clean-up, fast data overview, multiple statistical testing, differential expression analysis, phospho-site annotation and translation across species, multi-level enrichment analyses, proteome-wide kinase activity and substrate mapping and network hub analysis. Data output includes graphical formats such as heatmap, box-, volcano- and circos-plots. This resource is designed to assist proteome-wide data mining of pathophysiological mechanism without a need for programming knowledge.

bioinformatics↗

bioRxiv2021

Enhanced longitudinal differential expression detection in proteomics with robust reproducibility optimization regression

Quantitative proteomics has matured into an established tool and longitudinal proteomic experiments have begun to emerge. However, no effective, simple-to-use differential expression method for longitudinal proteomics data has been released. Typically, such data is noisy, contains missing values, has only few time points and biological replicates. To address this need, we provide a comprehensive evaluation of several existing differential expression methods for high-throughput longitudinal omics data and introduce a new method, Robust longitudinal Differential Expression (RolDE). The methods were evaluated using nearly 2000 semi-simulated spike-in proteomic datasets and a large experimental dataset. The RolDE method performed overall best; it was most tolerant to missing values, displayed good reproducibility and was the top method in ranking the results in a biologically meaningful way. Furthermore, contrary to many approaches, the open source RolDE does not require prior knowledge concerning the types of differences searched, but can easily be applied even by non-experienced users.

bioinformatics↗

bioRxiv2020

Metaproteomics boosted up by untargeted data-independent acquisition data analysis framework

Mass spectrometry based metaproteomics is a relatively new field of research that provides the ability to characterize the functionality of microbiota. Recently, we were the first to demonstrate the applicability of data-independent acquisition (DIA) mass spectrometry to the analysis of complex metaproteomic samples. This allowed us to circumvent many of the drawbacks of the conventionally used data-dependent acquisition (DDA) mass spectrometry, mainly the limited reproducibility when analyzing samples with complex microbial composition. However, the previous method still required additional DDA data on the samples to assist the DIA analysis. Here, we introduce, for the first time, a DIA metaproteomics approach that does not require any DDA data, but instead replaces a spectral library generated from DDA data with a pseudospectral library generated directly from the metaproteomics DIA samples. We demonstrate that using the new DIA-only approach, we can achieve higher peptide yields than with the DDA-assisted approach, while the amount of required mass spectrometry data is reduced to a single DIA run per sample. The new DIA-only metaproteomics approach is implemented as open-source software package DIAtools 2.0, which is freely available from DockerHub.

bioinformatics↗

bioRxiv2018

Data-independent acquisition mass spectrometry enables reproducible characterization of microbiota function

Metaproteomics is an emerging research area which aims to reveal the functionality of microbial communities - unlike the increasingly popular metagenomics providing insights only on the functional potential. So far, the common approach in metaproteomics has been data-dependent acquisition mass spectrometry (DDA). However, DDA is known to have limited reproducibility and dynamic range with samples of complex microbial composition. To overcome these limitations, we introduce here a novel approach utilizing data-independent acquisition (DIA) mass spectrometry, which has not been applied in metaproteomics of complex samples before. For robust analysis of the data, we introduce an open-source software package diatools, which is freely available at Docker Hub and runs on various operating systems. Our highly reproducible results on laboratory-assembled microbial mixtures and human fecal samples support the utility of our approach for functional characterization of complex microbiota. Hence, the approach is expected to dramatically improve our understanding on the role of microbiota in health and disease.

microbiology↗