Accessible and reproducible mass spectrometry imaging data analysis in Galaxy

Föll, Melanie Christine; Moritz, Lennart; Wollmann, Thomas; Stillger, Maren Nicole; Vockert, N.; Werner, Martin; Bronsert, Peter; Rohr, Karl; Gruening, Bjoern; Schilling, Oliver

doi:10.1101/628719

Cited by 6 publications

(6 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This code was (presumably) hand-written by the authors. The exception is the code in the training step 38 . This code seems to have been generated by the MAT-LAB Classification Learner app 39 .…”

Section: Cs Matlabmentioning

confidence: 99%

“…They discuss pre-processing the data through principal component analysis (PCA) and perform a comparison between multiple ML classification techniques. This 38 train RUSBoost Classifier 30l 1000s 20190222.m. 39 https://www.mathworks.com/help/stats/ classificationlearner-app.html knowledge of both domains is reflected by the classifications given by our framework, where all of the problem, solution workflow, and implementation are blended.…”

Section: Cs Matlabmentioning

confidence: 99%

“…For these particular workflows, the majority of the DS components are specific to MSI as they have been created by the authors in a previous work [38]. These components are wrappers around the Cardinal tool, which is a R language module specifically for analysing mass spectrometry-based imaging [9].…”

Section: Cs Galaxymentioning

confidence: 99%

See 2 more Smart Citations

Building Domain-Specific Machine Learning Workflows: A Conceptual Framework for the State-of-the-Practice

Oakes¹,

Famelis²,

Sahraoui³

2022

Preprint

View full text Add to dashboard Cite

Domain experts are increasingly employing machine learning to solve their domain-specific problems. This article presents six key challenges that a domain expert faces in transforming their problem into a computational workflow, and then into an executable implementation. These challenges arise out of our conceptual framework which presents the "route" of options that a domain expert may choose to take while developing their solution.To ground our conceptual framework in the stateof-the-practice, this article discusses a selection of available textual and graphical workflow systems and their support for these six challenges. Case studies from the literature in various domains are also examined to highlight the tools used by the domain experts as well as a classification of the domain-specificity and machine learning usage of their problem, workflow, and implementation.The state-of-the-practice informs our discussion of the six key challenges, where we identify which challenges are not sufficiently addressed by available tools. We also suggest possible research directions for software engineering researchers to increase the automation of these tools and disseminate best-practice techniques between software engineering and various scientific domains.

show abstract

Section: Cs Matlabmentioning

confidence: 99%

Section: Cs Matlabmentioning

confidence: 99%

See 1 more Smart Citation

Building Domain-Specific Machine Learning Workflows: A Conceptual Framework for the State-of-the-Practice

Oakes¹,

Famelis²,

Sahraoui³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Subregion annotations containing 3152 mass spectra in total and 77 spectra per tissue were extracted via an affine transformation strategy ( Föll et al , 2019 ). The two datasets were resampled, combined and pre-processed using Cardinal and MALDIquant algorithms on https://usegalaxy.eu ( Bemis et al , 2015 ; Föll et al , 2019 ; Gibb and Strimmer, 2012 ). The major pre-processing steps comprised peak picking, re-calibration, removal of contaminants and TIC normalization.…”

Section: Datamentioning

confidence: 99%

Deep multiple instance learning classifies subtissue locations in mass spectrometry images from tissue-level annotations

et al. 2020

Self Cite

View full text Add to dashboard Cite

Motivation Mass spectrometry imaging (MSI) characterizes the molecular composition of tissues at spatial resolution, and has a strong potential for distinguishing tissue types, or disease states. This can be achieved by supervised classification, which takes as input MSI spectra, and assigns class labels to subtissue locations. Unfortunately, developing such classifiers is hindered by the limited availability of training sets with subtissue labels as the ground truth. Subtissue labeling is prohibitively expensive, and only rough annotations of the entire tissues are typically available. Classifiers trained on data with approximate labels have sub-optimal performance. Results To alleviate this challenge, we contribute a semi-supervised approach mi-CNN. mi-CNN implements multiple instance learning with a convolutional neural network (CNN). The multiple instance aspect enables weak supervision from tissue-level annotations when classifying subtissue locations. The convolutional architecture of the CNN captures contextual dependencies between the spectral features. Evaluations on simulated and experimental datasets demonstrated that mi-CNN improved the subtissue classification as compared to traditional classifiers. We propose mi-CNN as an important step toward accurate subtissue classification in MSI, enabling rapid distinction between tissue types and disease states. Availability and implementation The data and code are available at https://github.com/Vitek-Lab/mi-CNN_MSI.

show abstract

“…Meanwhile, Cardinal was released in 2015, had a major update in 2019, and has even been partially adapted into a Galaxy workflow. 23 To utilize Cardinal, a certain level of coding knowledge in the R language as well as familiarity of open source mass spectrometry data formats is necessary as users may encounter situations that require troubleshooting. Although it may have a steep learning curve, Cardinal has the advantage of being a cost-friendly solution for data structures and workflows that can be used to incorporate new algorithms.…”

Section: ■ Conclusionmentioning

confidence: 99%

Evaluation of Data Analysis Platforms and Compatibility with MALDI-TOF Imaging Mass Spectrometry Data Sets

Luu

Condren

Kahl

et al. 2020

J. Am. Soc. Mass Spectrom.

View full text Add to dashboard Cite

Imaging mass spectrometry (IMS) has proven to be a useful tool when investigating the spatial distributions of metabolites and proteins in a biological system. One of the biggest advantages of IMS is the ability to maintain the 3D chemical composition of a sample and analyze it in a label-free manner. However, acquiring the spatial information leads to an increase in data size. Due to the increased availability of commercial mass spectrometers capable of IMS, there has been an exciting development of different statistical tools that can help decipher the spatial relevance of an analyte in a biological sample. To address this need, software packages like SCiLS and the open source R package Cardinal have been designed to perform unbiased spectral grouping based on the similarity of spectra in an IMS data set. In this note, we evaluate SCiLS and Cardinal compatibility with MALDI-TOF IMS data sets of the Gram-negative pathogen Pseudomonas aeruginosa PA14. Both software were able to perform unsupervised segmentation with similar performance. There were a few notable differences which are discussed related to the identification of statistically significant features which required optimization of preprocessing steps, region of interest, and manual analysis.

show abstract

Accessible and reproducible mass spectrometry imaging data analysis in Galaxy

Abstract: Background:

Cited by 6 publications

References 56 publications

Building Domain-Specific Machine Learning Workflows: A Conceptual Framework for the State-of-the-Practice

Building Domain-Specific Machine Learning Workflows: A Conceptual Framework for the State-of-the-Practice

Deep multiple instance learning classifies subtissue locations in mass spectrometry images from tissue-level annotations

Evaluation of Data Analysis Platforms and Compatibility with MALDI-TOF Imaging Mass Spectrometry Data Sets

Contact Info

Product

Resources

About