Osteoarthritis endotype discovery via clustering of biochemical marker data.

OBJECTIVES: Osteoarthritis (OA) patient stratification is an important challenge to design tailored treatments and drive drug development. Biochemical markers reflecting joint tissue turnover were measured in the IMI-APPROACH cohort at baseline and analysed using a machine learning approach in order to study OA-dominant phenotypes driven by the endotype-related clusters and discover the driving features and their disease-context meaning. METHOD: Data quality assessment was performed to design appropriate data preprocessing techniques. The k-means clustering algorithm was used to find dominant subgroups of patients based on the biochemical markers data. Classification models were trained to predict cluster membership, and Explainable AI techniques were used to interpret these to reveal the driving factors behind each cluster and identify phenotypes. Statistical analysis was performed to compare differences between clusters with respect to other markers in the IMI-APPROACH cohort and the longitudinal disease progression. RESULTS: Three dominant endotypes were found, associated with three phenotypes: C1) low tissue turnover (low repair and articular cartilage/subchondral bone turnover), C2) structural damage (high bone formation/resorption, cartilage degradation) and C3) systemic inflammation (joint tissue degradation, inflammation, cartilage degradation). The method achieved consistent results in the FNIH/OAI cohort. C1 had the highest proportion of non-progressors. C2 was mostly linked to longitudinal structural progression, and C3 was linked to sustained or progressive pain. CONCLUSIONS: This work supports the existence of differential phenotypes in OA. The biomarker approach could potentially drive stratification for OA clinical trials and contribute to precision medicine strategies for OA progression in the future. TRIAL REGISTRATION NUMBER: NCT03883568.

Predicted and actual 2-year structural and pain progression in the IMI-APPROACH knee osteoarthritis cohort.

OBJECTIVES: The IMI-APPROACH knee osteoarthritis study used machine learning (ML) to predict structural and/or pain progression, expressed by a structural (S) and pain (P) predicted-progression score, to select patients from existing cohorts. This study evaluates the actual 2-year progression within the IMI-APPROACH, in relation to the predicted-progression scores. METHODS: Actual structural progression was measured using minimum joint space width (minJSW). Actual pain (progression) was evaluated using the Knee injury and Osteoarthritis Outcomes Score (KOOS) pain questionnaire. Progression was presented as actual change (Δ) after 2 years, and as progression over 2 years based on a per patient fitted regression line using 0, 0.5, 1 and 2-year values. Differences in predicted-progression scores between actual progressors and non-progressors were evaluated. Receiver operating characteristic (ROC) curves were constructed and corresponding area under the curve (AUC) reported. Using Youden's index, optimal cut-offs were chosen to enable evaluation of both predicted-progression scores to identify actual progressors. RESULTS: Actual structural progressors were initially assigned higher S predicted-progression scores compared with structural non-progressors. Likewise, actual pain progressors were assigned higher P predicted-progression scores compared with pain non-progressors. The AUC-ROC for the S predicted-progression score to identify actual structural progressors was poor (0.612 and 0.599 for Δ and regression minJSW, respectively). The AUC-ROC for the P predicted-progression score to identify actual pain progressors were good (0.817 and 0.830 for Δ and regression KOOS pain, respectively). CONCLUSION: The S and P predicted-progression scores as provided by the ML models developed and used for the selection of IMI-APPROACH patients were to some degree able to distinguish between actual progressors and non-progressors. TRIAL REGISTRATION: ClinicalTrials.gov, https://clinicaltrials.gov, NCT03883568.

JEPETTO: a Cytoscape plugin for gene set enrichment and topological analysis based on interaction networks.

SUMMARY: JEPETTO (Java Enrichment of Pathways Extended To TOpology) is a Cytoscape 3.x plugin performing integrative human gene set analysis. It identifies functional associations between genes and known cellular pathways, and processes using protein interaction networks and topological analysis. The plugin integrates information from three separate web servers we published previously, specializing in enrichment analysis, pathways expansion and topological matching. This integration substantially simplifies the analysis of user gene sets and the interpretation of the results. We demonstrate the utility of the JEPETTO plugin on a set of misregulated genes associated with Alzheimer's disease. AVAILABILITY: Source code and binaries are freely available for download at http://apps.cytoscape.org/apps/jepetto, implemented in Java and multi-platform. Installable directly via Cytoscape plugin manager. Released under the GNU General Public Licence.

Transcriptional dynamics of two seed compartments with opposing roles in Arabidopsis seed germination.

Seed germination is a critical stage in the plant life cycle and the first step toward successful plant establishment. Therefore, understanding germination is of important ecological and agronomical relevance. Previous research revealed that different seed compartments (testa, endosperm, and embryo) control germination, but little is known about the underlying spatial and temporal transcriptome changes that lead to seed germination. We analyzed genome-wide expression in germinating Arabidopsis (Arabidopsis thaliana) seeds with both temporal and spatial detail and provide Web-accessible visualizations of the data reported (vseed.nottingham.ac.uk). We show the potential of this high-resolution data set for the construction of meaningful coexpression networks, which provide insight into the genetic control of germination. The data set reveals two transcriptional phases during germination that are separated by testa rupture. The first phase is marked by large transcriptome changes as the seed switches from a dry, quiescent state to a hydrated and active state. At the end of this first transcriptional phase, the number of differentially expressed genes between consecutive time points drops. This increases again at testa rupture, the start of the second transcriptional phase. Transcriptome data indicate a role for mechano-induced signaling at this stage and subsequently highlight the fates of the endosperm and radicle: senescence and growth, respectively. Finally, using a phylotranscriptomic approach, we show that expression levels of evolutionarily young genes drop during the first transcriptional phase and increase during the second phase. Evolutionarily old genes show an opposite pattern, suggesting a more conserved transcriptome prior to the completion of germination.

Contact map prediction using a large-scale ensemble of rule sets and the fusion of multiple predicted structural features.

MOTIVATION: The prediction of a protein's contact map has become in recent years, a crucial stepping stone for the prediction of the complete 3D structure of a protein. In this article, we describe a methodology for this problem that was shown to be successful in CASP8 and CASP9. The methodology is based on (i) the fusion of the prediction of a variety of structural aspects of protein residues, (ii) an ensemble strategy used to facilitate the training process and (iii) a rule-based machine learning system from which we can extract human-readable explanations of the predictor and derive useful information about the contact map representation. RESULTS: The main part of the evaluation is the comparison against the sequence-based contact prediction methods from CASP9, where our method presented the best rank in five out of the six evaluated metrics. We also assess the impact of the size of the ensemble used in our predictor to show the trade-off between performance and training time of our method. Finally, we also study the rule sets generated by our machine learning system. From this analysis, we are able to estimate the contribution of the attributes in our representation and how these interact to derive contact predictions. AVAILABILITY: http://icos.cs.nott.ac.uk/servers/psp.html. CONTACT: natalio.krasnogor@nottingham.ac.uk SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Development and validation of a machine learning-supported strategy of patient selection for osteoarthritis clinical trials: the IMI-APPROACH study.

Objectives: To efficiently assess the disease-modifying potential of new osteoarthritis treatments, clinical trials need progression-enriched patient populations. To assess whether the application of machine learning results in patient selection enrichment, we developed a machine learning recruitment strategy targeting progressive patients and validated it in the IMI-APPROACH knee osteoarthritis prospective study. Design: We designed a two-stage recruitment process supported by machine learning models trained to rank candidates by the likelihood of progression. First stage models used data from pre-existing cohorts to select patients for a screening visit. The second stage model used screening data to inform the final inclusion. The effectiveness of this process was evaluated using the actual 24-month progression. Results: From 3500 candidate patients, 433 with knee osteoarthritis were screened, 297 were enrolled, and 247 completed the 2-year follow-up visit. We observed progression related to pain (P, 30%), structure (S, 13%), and combined pain and structure (P â€‹+ â€‹S, 5%), and a proportion of non-progressors (N, 52%) ∼15% lower vs an unenriched population. Our model predicted these outcomes with AUC of 0.86 [95% CI, 0.81-0.90] for pain-related progression and AUC of 0.61 [95% CI, 0.52-0.70] for structure-related progression. Progressors were ranked higher than non-progressors for P â€‹+ â€‹S (median rank 65 vs 143, AUC = 0.75), P (median rank 77 vs 143, AUC = 0.71), and S patients (median rank 107 vs 143, AUC = 0.57). Conclusions: The machine learning-supported recruitment resulted in enriched selection of progressive patients. Further research is needed to improve structural progression prediction and assess this strategy in an interventional trial.

Baseline clinical characteristics of predicted structural and pain progressors in the IMI-APPROACH knee OA cohort.

OBJECTIVES: To describe the relations between baseline clinical characteristics of the Applied Public-Private Research enabling OsteoArthritis Clinical Headway (IMI-APPROACH) participants and their predicted probabilities for knee osteoarthritis (OA) structural (S) progression and/or pain (P) progression. METHODS: Baseline clinical characteristics of the IMI-APPROACH participants were used for this study. Radiographs were evaluated according to Kellgren and Lawrence (K&L grade) and Knee Image Digital Analysis. Knee Injury and Osteoarthritis Outcome Score (KOOS) and Numeric Rating Scale (NRS) were used to evaluate pain. Predicted progression scores for each individual were determined using machine learning models. Pearson correlation coefficients were used to evaluate correlations between scores for predicted progression and baseline characteristics. T-tests and χ2 tests were used to evaluate differences between participants with high versus low progression scores. RESULTS: Participants with high S progressions score were found to have statistically significantly less structural damage compared with participants with low S progression scores (minimum Joint Space Width, minJSW 3.56 mm vs 1.63 mm; p<0.001, K&L grade; p=0.028). Participants with high P progression scores had statistically significantly more pain compared with participants with low P progression scores (KOOS pain 51.71 vs 82.11; p<0.001, NRS pain 6.7 vs 2.4; p<0.001). CONCLUSIONS: The baseline minJSW of the IMI-APPROACH participants contradicts the idea that the (predicted) course of knee OA follows a pattern of inertia, where patients who have progressed previously are more likely to display further progression. In contrast, for pain progressors the pattern of inertia seems valid, since participants with high P score already have more pain at baseline compared with participants with a low P score.

Linking Engineered Cells to Their Digital Twins: A Version Control System for Strain Engineering.

As DNA sequencing and synthesis become cheaper and more easily accessible, the scale and complexity of biological engineering projects is set to grow. Yet, although there is an accelerating convergence between biotechnology and digital technology, a deficit in software and laboratory techniques diminishes the ability to make biotechnology more agile, reproducible, and transparent while, at the same time, limiting the security and safety of synthetic biology constructs. To partially address some of these problems, this paper presents an approach for physically linking engineered cells to their digital footprint-we called it digital twinning. This enables the tracking of the entire engineering history of a cell line in a specialized version control system for collaborative strain engineering via simple barcoding protocols.

Multi-classifier prediction of knee osteoarthritis progression from incomplete imbalanced longitudinal data.

Conventional inclusion criteria used in osteoarthritis clinical trials are not very effective in selecting patients who would benefit from a therapy being tested. Typically majority of selected patients show no or limited disease progression during a trial period. As a consequence, the effect of the tested treatment cannot be observed, and the efforts and resources invested in running the trial are not rewarded. This could be avoided, if selection criteria were more predictive of the future disease progression. In this article, we formulated the patient selection problem as a multi-class classification task, with classes based on clinically relevant measures of progression (over a time scale typical for clinical trials). Using data from two long-term knee osteoarthritis studies OAI and CHECK, we tested multiple algorithms and learning process configurations (including multi-classifier approaches, cost-sensitive learning, and feature selection), to identify the best performing machine learning models. We examined the behaviour of the best models, with respect to prediction errors and the impact of used features, to confirm their clinical relevance. We found that the model-based selection outperforms the conventional inclusion criteria, reducing by 20-25% the number of patients who show no progression. This result might lead to more efficient clinical trials.

Functional networks inference from rule-based machine learning models.

BACKGROUND: Functional networks play an important role in the analysis of biological processes and systems. The inference of these networks from high-throughput (-omics) data is an area of intense research. So far, the similarity-based inference paradigm (e.g. gene co-expression) has been the most popular approach. It assumes a functional relationship between genes which are expressed at similar levels across different samples. An alternative to this paradigm is the inference of relationships from the structure of machine learning models. These models are able to capture complex relationships between variables, that often are different/complementary to the similarity-based methods. RESULTS: We propose a protocol to infer functional networks from machine learning models, called FuNeL. It assumes, that genes used together within a rule-based machine learning model to classify the samples, might also be functionally related at a biological level. The protocol is first tested on synthetic datasets and then evaluated on a test suite of 8 real-world datasets related to human cancer. The networks inferred from the real-world data are compared against gene co-expression networks of equal size, generated with 3 different methods. The comparison is performed from two different points of view. We analyse the enriched biological terms in the set of network nodes and the relationships between known disease-associated genes in a context of the network topology. The comparison confirms both the biological relevance and the complementary character of the knowledge captured by the FuNeL networks in relation to similarity-based methods and demonstrates its potential to identify known disease associations as core elements of the network. Finally, using a prostate cancer dataset as a case study, we confirm that the biological knowledge captured by our method is relevant to the disease and consistent with the specialised literature and with an independent dataset not used in the inference process. AVAILABILITY: The implementation of our network inference protocol is available at: http://ico2s.org/software/funel.html.

Meta-stochastic simulation of biochemical models for systems and synthetic biology.

Stochastic simulation algorithms (SSAs) are used to trace realistic trajectories of biochemical systems at low species concentrations. As the complexity of modeled biosystems increases, it is important to select the best performing SSA. Numerous improvements to SSAs have been introduced but they each only tend to apply to a certain class of models. This makes it difficult for a systems or synthetic biologist to decide which algorithm to employ when confronted with a new model that requires simulation. In this paper, we demonstrate that it is possible to determine which algorithm is best suited to simulate a particular model and that this can be predicted a priori to algorithm execution. We present a Web based tool ssapredict that allows scientists to upload a biochemical model and obtain a prediction of the best performing SSA. Furthermore, ssapredict gives the user the option to download our high performance simulator ngss preconfigured to perform the simulation of the queried biochemical model with the predicted fastest algorithm as the simulation engine. The ssapredict Web application is available at http://ssapredict.ico2s.org. It is free software and its source code is distributed under the terms of the GNU Affero General Public License.

Hard Data Analytics Problems Make for Better Data Analysis Algorithms: Bioinformatics as an Example.

Data mining and knowledge discovery techniques have greatly progressed in the last decade. They are now able to handle larger and larger datasets, process heterogeneous information, integrate complex metadata, and extract and visualize new knowledge. Often these advances were driven by new challenges arising from real-world domains, with biology and biotechnology a prime source of diverse and hard (e.g., high volume, high throughput, high variety, and high noise) data analytics problems. The aim of this article is to show the broad spectrum of data mining tasks and challenges present in biological data, and how these challenges have driven us over the years to design new data mining and knowledge discovery procedures for biodata. This is illustrated with the help of two kinds of case studies. The first kind is focused on the field of protein structure prediction, where we have contributed in several areas: by designing, through regression, functions that can distinguish between good and bad models of a protein's predicted structure; by creating new measures to characterize aspects of a protein's structure associated with individual positions in a protein's sequence, measures containing information that might be useful for protein structure prediction; and by creating accurate estimators of these structural aspects. The second kind of case study is focused on omics data analytics, a class of biological data characterized for having extremely high dimensionalities. Our methods were able not only to generate very accurate classification models, but also to discover new biological knowledge that was later ratified by experimentalists. Finally, we describe several strategies to tightly integrate knowledge extraction and data mining in order to create a new class of biodata mining algorithms that can natively embrace the complexity of biological data, efficiently generate accurate information in the form of classification/regression models, and extract valuable new knowledge. Thus, a complete data-to-information-to-knowledge pipeline is presented.

Heuristic for maximizing DNA reuse in synthetic DNA library assembly.

De novo DNA synthesis is in need of new ideas for increasing production rate and reducing cost. DNA reuse in combinatorial library construction is one such idea. Here, we describe an algorithm for planning multistage assembly of DNA libraries with shared intermediates that greedily attempts to maximize DNA reuse, and show both theoretically and empirically that it runs in linear time. We compare solution quality and algorithmic performance to the best results reported for computing DNA assembly graphs, finding that our algorithm achieves solutions of equivalent quality but with dramatically shorter running times and substantially improved scalability. We also show that the related computational problem bounded-depth min-cost string production (BDMSP), which captures DNA library assembly operations with a simplified cost model, is NP-hard and APX-hard by reduction from vertex cover. The algorithm presented here provides solutions of near-minimal stages and thanks to almost instantaneous planning of DNA libraries it can be used as a metric of "manufacturability" to guide DNA library design. Rapid planning remains applicable even for DNA library sizes vastly exceeding today's biochemical assembly methods, future-proofing our method.