| 1. |
- Buric, Filip, 1988, et al.
(author)
-
Parallel Factor Analysis Enables Quantification and Identification of Highly Convolved Data-Independent-Acquired Protein Spectra
- 2020
-
In: Patterns. - : Elsevier BV. - 2666-3899. ; 1:9
-
Journal article (peer-reviewed)abstract
- The latest high-throughput mass spectrometry-based technologies can record virtually all molecules from complex biological samples, providing a holistic picture of proteomes in cells and tissues and enabling an evaluation of the overall status of a person's health. However, current best practices are still only scratching the surface of the wealth of available information obtained from the massive proteome datasets, and efficient novel data-driven strategies are needed. Powered by advances in GPU hardware and open-source machine-learning frameworks, we developed a data-driven approach, CANDIA, which disassembles highly complex proteomics data into the elementary molecular signatures of the proteins in biological samples. Our work provides a performant and adaptable solution that complements existing mass spectrometry techniques. As the central mathematical methods are generic, other scientific fields that are dealing with highly convolved datasets will benefit from this work.
|
|
| 2. |
- Calvanese, Diego, et al.
(author)
-
Accessing scientific data through knowledge graphs with Ontop
- 2021
-
In: Patterns. - : Elsevier. - 2666-3899. ; 2:10
-
Journal article (peer-reviewed)abstract
- Knowledge graphs (KGs) have recently gained attention due to their flexible data model, which reduces the effort needed for integration across different, possibly heterogeneous, data sources. In this tutorial, we learn how to access scientific data stored in a relational database through the virtual knowledge graph (VKG) approach. In such an approach, the data are exposed as a KG and enriched with semantic information coming from a domain ontology. The KG is “virtual” in the sense that the data are not replicated but stay within the data sources and are accessed at query time.We demonstrate the approach over scientific data coming from the biomedical domain and using the open-source VKG system Ontop. Since legacy data are exposed as a KG, users can access the data by means of a more convenient vocabulary provided by the domain ontology, benefit from automated reasoning capabilities, and do not need to focus on how the data are actually stored. Furthermore, the virtual approach allows for the use of KGs even in those contexts where the user does not own the data nor is granted the rights to make a copy of them.By relying on existing federation tools, the approach described here for accessing scientific data can also be used to integrate multiple, heterogeneous, and possibly semi-structured and unstructured data sources.Summary: In this tutorial, we learn how to set up and exploit the virtual knowledge graph (VKG) approach to access data stored in relational legacy systems and to enrich such data with domain knowledge coming from different heterogeneous (biomedical) resources. The VKG approach is based on an ontology that describes a domain of interest in terms of a vocabulary familiar to the user and exposes a high-level conceptual view of the data. Users can access the data by exploiting the conceptual view, and in this way they do not need to be aware of low-level storage details. They can easily integrate ontologies coming from different sources and can obtain richer answers thanks to the interaction between data and domain knowledge.
|
|
| 3. |
- Cope, Henry, et al.
(author)
-
Routine omics collection is a golden opportunity for European human research in space and analog environments
- 2022
-
In: PATTERNS. - : Elsevier BV. - 2666-3899. ; 3:10, s. 100550-
-
Journal article (peer-reviewed)abstract
- Widespread generation and analysis of omics data have revolutionized molecular medicine on Earth, yet its power to yield new mechanistic insights and improve occupational health during spaceflight is still to be fully realized in humans. Nevertheless, rapid technological advancements and ever-regular spaceflight programs mean that longitudinal, standardized, and cost-effective collection of human space omics data are firmly within reach. Here, we consider the practicality and scientific return of different sampling methods and omic types in the context of human spaceflight, We also appraise ethical and legal considerations pertinent to omics data derived from European astronauts and spaceflight participants (SFPs). Ultimately, we propose that a routine omics collection program in spaceflight and analog environments presents a golden opportunity. Unlocking this bright future of artificial intelligence (AI)-driven analyses and personalized medicine approaches will require further investigation into best practices, including policy design and standardization of omics data, metadata, and sampling methods.
|
|
| 4. |
- Gupta, Abhijit, et al.
(author)
-
Accurate prediction of B-form/A-form DNA conformation propensity from primary sequence : A machine learning and free energy handshake
- 2021
-
In: Patterns. - : Elsevier BV. - 2666-3899. ; 2:9
-
Journal article (peer-reviewed)abstract
- DNA carries the genetic code of life, with different conformations associated with different biological functions. Predicting the conformation of DNA from its primary sequence, although desirable, is a challenging problem owing to the polymorphic nature of DNA. We have deployed a host of machine learning algorithms, including the popular state-of-the-art LightGBM (a gradient boosting model), for building prediction models. We used the nested cross-validation strategy to address the issues of “overfitting” and selection bias. This simultaneously provides an unbiased estimate of the generalization performance of a machine learning algorithm and allows us to tune the hyperparameters optimally. Furthermore, we built a secondary model based on SHAP (SHapley Additive exPlanations) that offers crucial insight into model interpretability. Our detailed model-building strategy and robust statistical validation protocols tackle the formidable challenge of working on small datasets, which is often the case in biological and medical data.
|
|
| 5. |
- Hall, Ola, et al.
(author)
-
A review of explainable AI in the satellite data, deep machine learning, and human poverty domain
- 2022
-
In: Patterns. - Cambridge : Cell Press. - 2666-3899. ; 3:10
-
Research review (peer-reviewed)abstract
- Recent advances in artificial intelligence and deep machine learning have created a step change in how to measure human development indicators, in particular asset-based poverty. The combination of satellite imagery and deep machine learning now has the capability to estimate some types of poverty at a level close to what is achieved with traditional household surveys. An increasingly important issue beyond static estimations is whether this technology can contribute to scientific discovery and, consequently, new knowledge in the poverty and welfare domain. A foundation for achieving scientific insights is domain knowledge, which in turn translates into explainability and scientific consistency. We perform an integrative literature review focusing on three core elements relevant in this context—transparency, interpretability, and explainability—and investigate how they relate to the poverty, machine learning, and satellite imagery nexus. Our inclusion criteria for papers are that they cover poverty/wealth prediction, using survey data as the basis for the ground truth poverty/wealth estimates, be applicable to both urban and rural settings, use satellite images as the basis for at least some of the inputs (features), and the method should include deep neural networks. Our review of 32 papers shows that the status of the three core elements of explainable machine learning (transparency, interpretability, and domain knowledge) is varied and does not completely fulfill the requirements set up for scientific insights and discoveries. We argue that explainability is essential to support wider dissemination and acceptance of this research in the development community and that explainability means more than just interpretability. (c) 2022 The Author(s).
|
|
| 6. |
- Klasson, Marcus, et al.
(author)
-
Using Variational Multi-view Learning for Classification of Grocery Items
- 2020
-
In: Patterns. - : Elsevier. - 2666-3899. ; 1:8
-
Journal article (peer-reviewed)abstract
- An essential task for computer vision-based assistive technologies is to help visually impaired people to recognize objects in constrained environments, for instance, recognizing food items in grocery stores. In this paper, we introduce a novel dataset with natural images of groceries—fruits, vegetables, and packaged products—where all images have been taken inside grocery stores to resemble a shopping scenario. Additionally, we download iconic images and text descriptions for each item that can be utilized for better representation learning of groceries. We select a multi-view generative model, which can combine the different item information into lower-dimensional representations. The experiments show that utilizing the additional information yields higher accuracies on classifying grocery items than only using the natural images. We observe that iconic images help to construct representations separated by visual differences of the items, while text descriptions enable the model to distinguish between visually similar items by different ingredients.
|
|
| 7. |
- Lindberg, Staffan I., 1969, et al.
(author)
-
Data for Politics: Creating an International Research Infrastructure Measuring Democracy.
- 2020
-
In: Patterns. - : Elsevier BV. - 2666-3899. ; 1:4
-
Journal article (peer-reviewed)abstract
- Questions such as how democratic a country is, how free are its media, or how independent is its judiciary are highly important to researchers and decision makers. We describe a research infrastructure that produces the world's largest dataset on democracy, governance, human rights, and related topics. The dataset is far more resolved and accurate than previous efforts, currently covers 202 political units from 1789 until the present, and is regularly updated each spring. The infrastructure involves an online survey of over 3,000 experts from 180 countries. Survey design and advanced statistical techniques are crucial for assuring data validity. The infrastructure also provides reports and analyses based on the data and easy-to-use tools for exploring and graphing the data.
|
|
| 8. |
- Lv, Zhihan, Dr. 1984-, et al.
(author)
-
BlockNet : Beyond reliable spatial Digital Twins to Parallel Metaverse
- 2022
-
In: Patterns. - : Cell Press. - 2666-3899. ; 3:5
-
Journal article (peer-reviewed)abstract
- The development of Digital Twins has enabled them to be widely applied to various fields represented by intelligent manufacturing. A Metaverse, which is parallel to the physical world, needs mature and secure Digital Twins technology in addition to Parallel Intelligence to enable it to evolve autonomously. We propose that Blockchain combined with other areas does not simultaneously require all of the basic elements. We extract the immutable characteristics of Blockchain and propose a secure multidimensional data storage solution called BlockNet that can ensure the security of the digital mapping process of the Internet of Things, thereby improving the data reliability of Digital Twins. Additionally, to address some of the challenges faced by multiscale spatial data processing, we propose a nonmutagenic multidimensional Hash Geocoding method, allowing unique indexing of multidimensional information and avoiding information loss due to data dimensionality reduction while improving the efficiency of information retrieval and facilitating the implementation of the Metaverse through spatial Digital Twins based on these two studies. © 2022 The Author(s)
|
|
| 9. |
- Nasimian, Ahmad, et al.
(author)
-
AlphaML : A clear, legible, explainable, transparent, and elucidative binary classification platform for tabular data
- 2024
-
In: Patterns. - 2666-3899. ; 5:1
-
Journal article (peer-reviewed)abstract
- Leveraging the potential of machine learning and recognizing the broad applications of binary classification, it becomes essential to develop platforms that are not only powerful but also transparent, interpretable, and user friendly. We introduce alphaML, a user-friendly platform that provides clear, legible, explainable, transparent, and elucidative (CLETE) binary classification models with comprehensive customization options. AlphaML offers feature selection, hyperparameter search, sampling, and normalization methods, along with 15 machine learning algorithms with global and local interpretation. We have integrated a custom metric for hyperparameter search that considers both training and validation scores, safeguarding against under- or overfitting. Additionally, we employ the NegLog2RMSL scoring method, which uses both training and test scores for a thorough model evaluation. The platform has been tested using datasets from multiple domains and offers a graphical interface, removing the need for programming expertise. Consequently, alphaML exhibits versatility, demonstrating promising applicability across a broad spectrum of tabular data configurations.
|
|
| 10. |
- Nicholas, Kimberly A., et al.
(author)
-
A harmonized and spatially explicit dataset from 16 million payments from the European Union's Common Agricultural Policy for 2015
- 2021
-
In: Patterns. - : Elsevier BV. - 2666-3899. ; 2:4
-
Journal article (peer-reviewed)abstract
- The Common Agricultural Policy (CAP) is the largest budget item in the European Union, but varied data reporting hampers holistic analysis. Here we have assembled the first dataset to our knowledge to report individual CAP payments by standardized CAP funding measures and geolocation. We created this dataset by translating, geolocating to the county or province (NUTS3) level, and consistently harmonizing payment measures for over 16 million payments from 2015, originally reported by EU member states and compiled by the Open Knowledge Foundation Germany. This dataset and code allow in-depth analysis of over €60 billion in public spending by purpose and location for the first time, which enables both individual payment tracing and analysis by aggregation. These data are representative of the distribution of annual CAP payments from 2014 to 2020 and are of interest to researchers, policy makers, non-governmental organizations, and journalists for evaluating the distribution and impacts of CAP spending.
|
|
