@article{ConradGenzelCvetkovicetal.2017, author = {Conrad, Tim and Genzel, Martin and Cvetkovic, Nada and Wulkow, Niklas and Vybiral, Jan and Kutyniok, Gitta and Sch{\"u}tte, Christof}, title = {Sparse Proteomics Analysis - a compressed sensing-based approach for feature selection and classification of high-dimensional proteomics mass spectrometry data}, volume = {18}, journal = {BMC Bioinformatics}, number = {160}, doi = {10.1186/s12859-017-1565-4}, pages = {1 -- 20}, year = {2017}, abstract = {Motivation: High-throughput proteomics techniques, such as mass spectrometry (MS)-based approaches, produce very high-dimensional data-sets. In a clinical setting one is often interested how MS spectra dier between patients of different classes, for example spectra from healthy patients vs. spectra from patients having a particular disease. Machine learning algorithms are needed to (a) identify these discriminating features and (b) classify unknown spectra based on this feature set. Since the acquired data is usually noisy, the algorithms should be robust to noise and outliers, and the identied feature set should be as small as possible. Results: We present a new algorithm, Sparse Proteomics Analysis (SPA), based on the theory of Compressed Sensing that allows to identify a minimal discriminating set of features from mass spectrometry data-sets. We show how our method performs on artificial and real-world data-sets.}, language = {en} } @article{SeeberConradHoppeetal.2017, author = {Seeber, L. and Conrad, Tim and Hoppe, Christian and Obermeier, Patrick and Chen, X. and Karsch, K. and Muehlhans, S. and Tief, Franziska and Boettcher, Sindy and Diedrich, S. and Schweiger, Brunhilde and Rath, Barbara}, title = {Educating parents about the vaccination status of their children: A user-centered mobile application}, volume = {5}, journal = {Preventive Medicine Reports}, doi = {10.1016/j.pmedr.2017.01.002}, pages = {241 -- 250}, year = {2017}, abstract = {Parents are often uncertain about the vaccination status of their children. In times of vaccine hesitancy, vaccination programs could benefit from active patient participation. The Vaccination App (VAccApp) was developed by the Vienna Vaccine Safety Initiative, enabling parents to learn about the vaccination status of their children, including 25 different routine, special indication and travel vaccines listed in the WHO Immunization Certificate of Vaccination (WHO-ICV). Between 2012 and 2014, the VAccApp was validated in a hospital-based quality management program in Berlin, Germany, in collaboration with the Robert Koch Institute. Parents of 178 children were asked to transfer the immunization data of their children from the WHO-ICV into the VAccApp. The respective WHO-ICV was photocopied for independent, professional data entry (gold standard). Demonstrating the status quo in vaccine information reporting, a Recall Group of 278 parents underwent structured interviews for verbal immunization histories, without the respective WHO-ICV. Only 9\% of the Recall Group were able to provide a complete vaccination status; on average 39\% of the questions were answered correctly. Using the WHO-ICV with the help of the VAccApp resulted in 62\% of parents providing a complete vaccination status; on average 95\% of the questions were answered correctly. After using the VAccApp, parents were more likely to remember key aspects of the vaccination history. User-friendly mobile applications empower parents to take a closer look at the vaccination record, thereby taking an active role in providing accurate vaccination histories. Parents may become motivated to ask informed questions and to keep vaccinations up-to-date.}, language = {en} } @article{ObermeierHeimBiereetal.2018, author = {Obermeier, Patrick and Heim, A. and Biere, Barbara and Hage, E. and Alchikh, Maren and Conrad, Tim and Schweiger, Brunhilde and Rath, Barbara}, title = {Clinical characteristics and disease severity associated with adenovirus infections in infants and children - discovery of a novel adenovirus, HAdV-D80}, journal = {Clinical Infectious Diseases}, year = {2018}, language = {en} } @article{KorkSpiesConradetal.2018, author = {Kork, F. and Spies, Claudia and Conrad, Tim and Weiss, B. and Roenneberg, T. and Wernecke, K.-D. and Balzer, Felix}, title = {Associations of postoperative mortality with the time of day, week and year}, journal = {Anaesthesia}, year = {2018}, language = {en} } @article{MirelesConrad2018, author = {Mireles, Victor and Conrad, Tim}, title = {Reusable building blocks in biological systems}, volume = {15}, journal = {Journal of the Royal Society Interface}, number = {149}, doi = {10.1098/rsif.2018.0595}, pages = {1 -- 10}, year = {2018}, abstract = {One of the most widely recognized features of biological systems is their modularity. The modules that constitute biological systems are said to be redeployed and combined across several conditions, thus acting as building blocks. In this work, we analyse to what extent are these building blocks reusable as compared with those found in randomized versions of a system. We develop a notion of decompositions of systems into phenotypic building blocks, which allows them to overlap while maximizing the number of times a building block is reused across several conditions. Different biological systems present building blocks whose reusability ranges from single use (e.g. condition specific) to constitutive, although their average reusability is not always higher than random equivalents of the system. These decompositions reveal a distinct distribution of building block sizes in real biological systems. This distribution stems, in part, from the peculiar usage pattern of the elements of biological systems, and constitutes a new angle to study the evolution of modularity.}, language = {en} } @article{MaConradAlchikhetal.2018, author = {Ma, Xiaolin and Conrad, Tim and Alchikh, Maren and Reiche, J. and Schweiger, Brunhilde and Rath, Barbara}, title = {Can we distinguish respiratory viral infections based on clinical features? A prospective pediatric cohort compared to systematic literature review}, volume = {28}, journal = {Medical Virology}, number = {5}, doi = {10.1002/rmv.1997}, pages = {1 -- 12}, year = {2018}, abstract = {Studies have shown that the predictive value of "clinical diagnoses" of influenza and other respiratory viral infections is low, especially in children. In routine care, pediatricians often resort to clinical diagnoses, even in the absence of robust evidence-based criteria. We used a dual approach to identify clinical characteristics that may help to differentiate infections with common pathogens including influenza, respiratory syncytial virus, adenovirus, metapneumovirus, rhinovirus, bocavirus-1, coronaviruses, or parainfluenza virus: (a) systematic review and meta-analysis of 47 clinical studies published in Medline (June 1996 to March 2017, PROSPERO registration number: CRD42017059557) comprising 49 858 individuals and (b) data-driven analysis of an inception cohort of 6073 children with ILI (aged 0-18 years, 56\% male, December 2009 to March 2015) examined at the point of care in addition to blinded PCR testing. We determined pooled odds ratios for the literature analysis and compared these to odds ratios based on the clinical cohort dataset. This combined analysis suggested significant associations between influenza and fever or headache, as well as between respiratory syncytial virus infection and cough, dyspnea, and wheezing. Similarly, literature and cohort data agreed on significant associations between HMPV infection and cough, as well as adenovirus infection and fever. Importantly, none of the abovementioned features were unique to any particular pathogen but were also observed in association with other respiratory viruses. In summary, our "real-world" dataset confirmed published literature trends, but no individual feature allows any particular type of viral infection to be ruled in or ruled out. For the time being, laboratory confirmation remains essential. More research is needed to develop scientifically validated decision models to inform best practice guidelines and targeted diagnostic algorithms.}, language = {en} } @article{AlchikhConradHoppeetal.2018, author = {Alchikh, Maren and Conrad, Tim and Hoppe, Christian and Ma, Xiaolin and Broberg, Eeva K. and Penttinen, P. and Reiche, J. and Biere, Barbara and Schweiger, Brunhilde and Rath, Barbara}, title = {Are we missing respiratory viral infections in infants and children? Comparison of a hospital-based quality management system with standard of care. Clinical Microbiology and Infection}, journal = {Clinical Microbiology and Infection}, number = {06/18}, doi = {10.1016/j.cmi.2018.05.023}, pages = {1 -- 1}, year = {2018}, language = {en} } @inproceedings{JayrannejadConrad2017, author = {Jayrannejad, Fahrnaz and Conrad, Tim}, title = {Better Interpretable Models for Proteomics Data Analysis Using rule-based Mining}, booktitle = {Springer Lecture Notes in Artificial Intelligence}, pages = {studi}, year = {2017}, abstract = {Recent advances in -omics technology has yielded in large data-sets in many areas of biology, such as mass spectrometry based proteomics. However, analyzing this data is still a challenging task mainly due to the very high dimensionality and high noise content of the data. One of the main objectives of the analysis is the identification of relevant patterns (or features) which can be used for classification of new samples to healthy or diseased. So, a method is required to find easily interpretable models from this data. To gain the above mentioned goal, we have adapted the disjunctive association rule mining algorithm, TitanicOR, to identify emerging patterns from our mass spectrometry proteomics data-sets. Comparison to five state-of-the-art methods shows that our method is advantageous them in terms of identifying the inter-dependency between the features and the TP-rate and precision of the features selected. We further demonstrate the applicability of our algorithm to one previously published clinical data-set.}, language = {en} } @article{ShaoBjaanaesHellandetal.2019, author = {Shao, Borong and Bjaanaes, Maria and Helland, Aslaug and Sch{\"u}tte, Christof and Conrad, Tim}, title = {EMT network-based feature selection improves prognosis prediction in lung adenocarcinoma}, volume = {14}, journal = {PLOS ONE}, number = {1}, doi = {10.1371/journal.pone.0204186}, year = {2019}, abstract = {Various feature selection algorithms have been proposed to identify cancer prognostic biomarkers. In recent years, however, their reproducibility is criticized. The performance of feature selection algorithms is shown to be affected by the datasets, underlying networks and evaluation metrics. One of the causes is the curse of dimensionality, which makes it hard to select the features that generalize well on independent data. Even the integration of biological networks does not mitigate this issue because the networks are large and many of their components are not relevant for the phenotype of interest. With the availability of multi-omics data, integrative approaches are being developed to build more robust predictive models. In this scenario, the higher data dimensions create greater challenges. We proposed a phenotype relevant network-based feature selection (PRNFS) framework and demonstrated its advantages in lung cancer prognosis prediction. We constructed cancer prognosis relevant networks based on epithelial mesenchymal transition (EMT) and integrated them with different types of omics data for feature selection. With less than 2.5\% of the total dimensionality, we obtained EMT prognostic signatures that achieved remarkable prediction performance (average AUC values above 0.8), very significant sample stratifications, and meaningful biological interpretations. In addition to finding EMT signatures from different omics data levels, we combined these single-omics signatures into multi-omics signatures, which improved sample stratifications significantly. Both single- and multi-omics EMT signatures were tested on independent multi-omics lung cancer datasets and significant sample stratifications were obtained.}, language = {en} } @article{SarichDjurdjevacConradBruckneretal.2014, author = {Sarich, Marco and Djurdjevac Conrad, Natasa and Bruckner, Sharon and Conrad, Tim and Sch{\"u}tte, Christof}, title = {Modularity revisited: A novel dynamics-based concept for decomposing complex networks}, volume = {1}, journal = {Journal of Computational Dynamics}, number = {1}, doi = {10.3934/jcd.2014.1.191}, pages = {191 -- 212}, year = {2014}, language = {en} } @article{AlchikhConradMaetal.2019, author = {Alchikh, Maren and Conrad, Tim and Ma, Xiaolin and Broberg, Eeva K. and Penttinen, P. and Reiche, J. and Biere, Barbara and Schweiger, Brunhilde and Rath, Barbara and Hoppe, Christian}, title = {Are we missing respiratory viral infections in infants and children? Comparison of a hospital-based quality management system with standard of care}, volume = {25}, journal = {Clinical Microbiology and Infection}, number = {3}, issn = {1469-0691}, doi = {10.1016/j.cmi.2018.05.023}, pages = {380.e9 -- 380.e16}, year = {2019}, language = {en} } @article{RamsConrad2020, author = {Rams, Mona and Conrad, Tim}, title = {Dictionary Learning for transcriptomics data reveals type-specific gene modules in a multi-class setting}, volume = {62}, journal = {it - Information Technology}, number = {3-4}, publisher = {De Gruyter}, address = {Oldenbourg}, issn = {2196-7032}, doi = {https://doi.org/10.1515/itit-2019-0048}, year = {2020}, language = {en} } @article{RathConradMylesetal.2017, author = {Rath, Barbara and Conrad, Tim and Myles, Puja and Alchikh, Maren and Ma, Xiaolin and Hoppe, Christian and Tief, Franziska and Chen, Xi and Obermeier, Patrick and Kisler, Bron and Schweiger, Brunhilde}, title = {Influenza and other respiratory viruses: standardizing disease severity in surveillance and clinical trials}, volume = {15}, journal = {Expert Review of Anti-infective Therapy}, number = {6}, doi = {10.1080/14787210.2017.1295847}, pages = {545 -- 568}, year = {2017}, abstract = {Introduction: Influenza-Like Illness is a leading cause of hospitalization in children. Disease burden due to influenza and other respiratory viral infections is reported on a population level, but clinical scores measuring individual changes in disease severity are urgently needed. Areas covered: We present a composite clinical score allowing individual patient data analyses of disease severity based on systematic literature review and WHO-criteria for uncomplicated and complicated disease. The 22-item ViVI Disease Severity Score showed a normal distribution in a pediatric cohort of 6073 children aged 0-18 years (mean age 3.13; S.D. 3.89; range: 0 to 18.79). Expert commentary: The ViVI Score was correlated with risk of antibiotic use as well as need for hospitalization and intensive care. The ViVI Score was used to track children with influenza, respiratory syncytial virus, human metapneumovirus, human rhinovirus, and adenovirus infections and is fully compliant with regulatory data standards. The ViVI Disease Severity Score mobile application allows physicians to measure disease severity at the point-of care thereby taking clinical trials to the next level.}, language = {en} } @article{ShaoCannistraciConrad2017, author = {Shao, Borong and Cannistraci, Carlo Vittorio and Conrad, Tim}, title = {Epithelial Mesenchymal Transition Network-Based Feature Engineering in Lung Adenocarcinoma Prognosis Prediction Using Multiple Omic Data}, volume = {3}, journal = {Genomics and Computational Biology}, number = {3}, doi = {10.18547/gcb.2017.vol3.iss3.e57}, pages = {e57}, year = {2017}, abstract = {Epithelial mesenchymal transition (EMT) process has been shown as highly relevant to cancer prognosis. However, although different biological network-based biomarker identification methods have been proposed to predict cancer prognosis, EMT network has not been directly used for this purpose. In this study, we constructed an EMT regulatory network consisting of 87 molecules and tried to select features that are useful for prognosis prediction in Lung Adenocarcinoma (LUAD). To incorporate multiple molecular profiles, we obtained four types of molecular data including mRNA-Seq, copy number alteration (CNA), DNA methylation, and miRNA-Seq data from The Cancer Genome Atlas. The data were mapped to the EMT network in three alternative ways: mRNA-Seq and miRNA-Seq, DNA methylation, and CNA and miRNA-Seq. Each mapping was employed to extract five different sets of features using discretization and network-based biomarker identification methods. Each feature set was then used to predict prognosis with SVM and logistic regression classifiers. We measured the prediction accuracy with AUC and AUPR values using 10 times 10-fold cross validation. For a more comprehensive evaluation, we also measured the prediction accuracies of clinical features, EMT plus clinical features, randomly picked 87 molecules from each data mapping, and using all molecules from each data type. Counter-intuitively, EMT features do not always outperform randomly selected features and the prediction accuracies of the five feature sets are mostly not significantly different. Clinical features are shown to give the highest prediction accuracies. In addition, the prediction accuracies of both EMT features and random features are comparable as using all features (more than 17,000) from each data type.}, language = {en} } @article{JayrannejadConrad2017, author = {Jayrannejad, Fahrnaz and Conrad, Tim}, title = {Better Interpretable Models for Proteomics Data Analysis Using rule-based Mining}, journal = {Springer Lecture Notes in Artificial Intelligence}, year = {2017}, abstract = {Recent advances in -omics technology has yielded in large data-sets in many areas of biology, such as mass spectrometry based proteomics. However, analyzing this data is still a challenging task mainly due to the very high dimensionality and high noise content of the data. One of the main objectives of the analysis is the identification of relevant patterns (or features) which can be used for classification of new samples to healthy or diseased. So, a method is required to find easily interpretable models from this data. To gain the above mentioned goal, we have adapted the disjunctive association rule mining algorithm, TitanicOR, to identify emerging patterns from our mass spectrometry proteomics data-sets. Comparison to five state-of-the-art methods shows that our method is advantageous them in terms of identifying the inter-dependency between the features and the TP-rate and precision of the features selected. We further demonstrate the applicability of our algorithm to one previously published clinical data-set.}, language = {en} } @article{MarzbanConradMarbanetal.2018, author = {Marzban, Forough and Conrad, Tim and Marban, Pouria and Sodoudi, Sahar}, title = {Estimation of the Near-Surface Air Temperature during the Day and Nighttime from MODIS in Berlin, Germany}, volume = {7}, journal = {International Journal of Advanced Remote Sensing and GIS}, number = {1}, doi = {10.23953/cloud.ijarsg.337}, pages = {2478 -- 2517}, year = {2018}, abstract = {Air temperature (Tair or T2m) is an important climatological variable for forest biosphere processes and climate change research. Due to the low density and the uneven distribution of weather stations, traditional ground-based observations cannot accurately capture the spatial distribution of Tair. In this study, Tair in Berlin is estimated during the day and night time over six land cover/land use (LC/LU) types by satellite remote sensing data over a large domain and a relatively long period (7 years). Aqua and Terra MODIS (Moderate Resolution Imaging Spectroradiometer) data and meteorological data for the period from 2007 to 2013 were collected to estimate Tair. Twelve environmental variables (land surface temperature (LST), normalized difference vegetation index (NDVI), Julian day, latitude, longitude, Emissivity31, Emissivity32, altitude, albedo, wind speed, wind direction and air pressure) were selected as predictors. Moreover, a comparison between LST from MODIS Terra and Aqua with daytime and night time air temperatures (Tday, Tnight) was done respectively and in addition, the spatial variability of LST and Tair relationship by applying a varying window size on the MODIS LST grid was examined. An analysis of the relationship between the observed Tair and the spatially averaged remotely sensed LST, indicated that 3 × 3 and 1 × 1 pixel size was the optimal window size for the statistical model estimating Tair from MODIS data during the day and night time, respectively. Three supervised learning methods (Adaptive Neuro Fuzzy Inference system (ANFIS), Artificial Neural Network (ANN) and Support vector machine (SVR)) were used to estimate Tair during the day and night time, and their performances were validated by cross-validation for each LC/LU. Moreover, tuning the hyper parameters of some models like SVR and ANN were investigated. For tuning the hyper parameters of SVR, Simulated Annealing (SA) was applied (SA-SVR model) and a multiple-layer feed-forward (MLF) neural networks with three layers and different nodes in hidden layers are used with Levenber-Marquardt back-propagation (LM-BP), in order to achieve higher accuracy in the estimation of Tair. Results indicated that the ANN model achieved better accuracy (RMSE= 2.16°C, MAE = 1.69°C, R2 = 0.95) than SA_SVR model (RMSE= 2.50°C, MAE = 1.92°C, R2 = 0.91) and ANFIS model (RMSE= 2.88°C, MAE= 2.2°C, R2 = 0.89) over six LC/LU during the day and night time. The Q-Q diagram of SA-SVR, ANFIS and NN show that all three models slightly tended to underestimate and overestimate the extreme and low temperatures for all LC/LU classes during the day and night time. The weak performance in the extreme and low temperatures are a consequence of the small number of data in these temperatures. These satisfactory results indicate that this approach is proper for estimating air temperature and spatial window size is an important factor that should be considered in the estimation of air temperature.}, language = {en} } @article{KorkConradWeissetal.2018, author = {Kork, F. and Conrad, Tim and Weiss, B. and Roenneberg, T. and Wernecke, K.-D. and Balzer, Felix}, title = {Associations of postoperative mortality with the time of day, week and year}, volume = {73}, journal = {Anaesthesia}, number = {6}, issn = {0959-2962}, pages = {711 -- 718}, year = {2018}, language = {en} } @article{ConradKarschObermeieretal.2015, author = {Conrad, Tim and Karsch, K. and Obermeier, Patrick and Seeber, L. and Chen, X. and Tief, Franziska and Muehlhans, S. and Hoppe, Christian and Boettcher, Sindy and Diedrich, S. and Rath, Barbara}, title = {Human Parechovirus Infections Associated with Seizures and Rash: A Syndromic Surveillance Study in Children}, volume = {34}, journal = {The Pediatric Infectious Disease Journal}, number = {10}, year = {2015}, language = {en} } @article{ConradLenga2015, author = {Conrad, Tim and Lenga, Matthias}, title = {Non-convex regularization for supervised learning}, journal = {Technical Report}, year = {2015}, language = {en} } @article{ConradMireles2015, author = {Conrad, Tim and Mireles, Victor}, title = {Minimum-overlap clusterings and the sparsity of overcomplete decompositions of binary matrics}, volume = {51}, journal = {Procedia Computer Science}, doi = {10.1016/j.procs.2015.05.500}, pages = {2967 -- 2971}, year = {2015}, abstract = {Given a set of n binary data points, a widely used technique is to group its features into k clusters. In the case where n {\ensuremath{<}} k, the question of how overlapping are the clusters becomes of interest. In this paper we approach the question through matrix decomposition, and relate the degree of overlap with the sparsity of one of the resulting matrices. We present analytical results regarding bounds on this sparsity, and a heuristic to estimate the minimum amount of overlap that an exact grouping of features into k clusters must have. As shown below, adding new data will not alter this minimum amount of overlap.}, language = {en} } @article{ConradShao2015, author = {Conrad, Tim and Shao, Borong}, title = {Are NoSQL data stores useful for bioinformatics researchers?}, volume = {3}, journal = {International Journal on Recent and Innovation Trends in Computing and Communication (IJRITCC)}, number = {3}, doi = {10.17762/ijritcc2321-8169.1503176}, pages = {1704 -- 1708}, year = {2015}, language = {en} } @article{ConradWallmeyerRamsetal.2015, author = {Conrad, Tim and Wallmeyer, Leonie and Rams, Mona and B{\"o}rno, Stefan and Timmermann, Bernd and Hedtrich, Sarah}, title = {Analyzing the Influence of Filaggrin on Other Skin Associated Genes in a Filaggrin Deficient Skin Model Using the Core Network Approach}, journal = {Technical Report}, year = {2015}, language = {en} } @article{ConradLeichtleCeglareketal.2013, author = {Conrad, Tim and Leichtle, Alexander Benedikt and Ceglarek, Uta and Weinert, P. and Nakas, C.T. and Nuoffer, Jean-Marc and Kase, Julia and Witzigmann, Helmut and Thiery, Joachim and Fiedler, Georg Martin}, title = {Pancreatic carcinoma, pancreatitis, and healthy controls - metabolite models in a three-class diagnostic dilemma}, volume = {9}, journal = {Metabolomics}, number = {3}, doi = {10.1007/s11306-012-0476-7}, pages = {677 -- 687}, year = {2013}, abstract = {Background: Metabolomics as one of the most rapidly growing technologies in the ?-omics?field denotes the comprehensive analysis of low molecular-weight compounds and their pathways. Cancer-specific alterations of the metabolome can be detected by high-throughput massspectrometric metabolite profiling and serve as a considerable source of new markers for the early differentiation of malignant diseases as well as their distinction from benign states. However, a comprehensive framework for the statistical evaluation of marker panels in a multi-class setting has not yet been established. Methods: We collected serum samples of 40 pancreatic carcinoma patients, 40 controls, and 23 pancreatitis patients according to standard protocols and generated amino acid profiles by routine mass-spectrometry. In an intrinsic three-class bioinformatic approach we compared these profiles, evaluated their selectivity and computed multi-marker panels combined with the conventional tumor marker CA 19-9. Additionally, we tested for non-inferiority and superiority to determine the diagnostic surplus value of our multi-metabolite marker panels.  Results: Compared to CA 19-9 alone, the combined amino acid-based metabolite panel had a superior selectivity for the discrimination of healthy controls, pancreatitis, and pancreatic carcinoma patients [Volume under ROC surface (VUS) = 0.891 (95\\% CI 0.794 - 0.968)]. Conclusions: We combined highly standardized samples, a three-class study design, a highthroughput mass-spectrometric technique, and a comprehensive bioinformatic framework to identify metabolite panels selective for all three groups in a single approach. Our results suggest that metabolomic profiling necessitates appropriate evaluation strategies and ?despite all its current limitations? can deliver marker panels with high selectivity even in multi-class settings.}, language = {en} } @article{GuptaReinschSpoetteretal.2013, author = {Gupta, Pooja and Reinsch, Norbert and Sp{\"o}tter, Andreas and Conrad, Tim and Bienefeld, Kaspar}, title = {Accuracy of the unified approach in maternally influenced traits - illustrated by a simulation study in the honey bee (Apis mellifera)}, volume = {14}, journal = {BMC Genetics}, number = {36}, doi = {10.1186/1471-2156-14-36}, year = {2013}, language = {en} } @article{ConradYou2016, author = {Conrad, Tim and You, Xintian}, title = {Acfs: accurate circRNA identification and quantification from NGS data}, volume = {6}, journal = {Nature Scientific Reports}, doi = {10.1038/srep38820}, year = {2016}, abstract = {Circular RNAs (circRNAs) are a group of single-stranded RNAs in closed circular form. They are splicing-generated, widely expressed in various tissues and have functional implications in development and diseases. To facilitate genome-wide characterization of circRNAs using RNA-Seq data, we present a freely available software package named acfs. Acfs allows de novo, accurate and fast identification and abundance quantification of circRNAs from single- and paired-ended RNA-Seq data. On simulated datasets, acfs achieved the highest F1 accuracy and lowest false discovery rate among current state-of-the-art tools. On real-world datasets, acfs efficiently identified more bona fide circRNAs. Furthermore, we demonstrated the power of circRNA analysis on two leukemia datasets. We identified a set of circRNAs that are differentially expressed between AML and APL samples, which might shed light on the potential molecular classification of complex diseases using circRNA profiles. Moreover, chromosomal translocation, as manifested in numerous diseases, could produce not only fusion transcripts but also fusion circRNAs of clinical relevance. Featured with high accuracy, low FDR and the ability to identify fusion circRNAs, we believe that acfs is well suited for a wide spectrum of applications in characterizing the landscape of circRNAs from non-model organisms to cancer biology.}, language = {en} } @article{MirelesConrad2016, author = {Mireles, Victor and Conrad, Tim}, title = {Decomposing biological systems into reusable modules reveals characteristic module size distributions}, year = {2016}, abstract = {One of the widely recognized features of biological systems is their modularity. The modules that comprise biological systems are said to be redeployed and combined across several conditions. In this work, we analyze to what extent are these modules indeed reusable as compared to randomized versions of a system. We develop a notion of modular decompositions of systems that allows for modules to overlap while maximizing the number of times a module is reused across several conditions. Different biological systems present modules whose reusability ranges from the condition specific to the constitutive, although their average reusability is not always higher than random equivalents of the system. These decompositions reveal a distinct distribution of module sizes in real biological systems. This distribution stems, in part, from the peculiar usage pattern of the elements of biological systems, and constitutes a new angle to study the evolution of modularity.}, language = {en} } @article{HoppeObermeierMehlhansetal.2016, author = {Hoppe, Christian and Obermeier, Patrick and Mehlhans, S. and Alchikh, Maren and Seeber, L. and Tief, Franziska and Karsch, K. and Chen, X. and Boettcher, Sindy and Diedrich, S. and Conrad, Tim}, title = {Innovative Digital Tools and Surveillance Systems for the Timely Detection of Adverse Events at the Point of Care: A Proof-of-Concept Study}, volume = {39}, journal = {Drug Safety}, number = {10}, doi = {10.1007/s40264-016-0437-6}, pages = {977 -- 988}, year = {2016}, abstract = {Regulatory authorities often receive poorly structured safety reports requiring considerable effort to investigate potential adverse events post hoc. Automated question-and-answer systems may help to improve the overall quality of safety information transmitted to pharmacovigilance agencies. This paper explores the use of the VACC-Tool (ViVI Automated Case Classification Tool) 2.0, a mobile application enabling physicians to classify clinical cases according to 14 pre-defined case definitions for neuroinflammatory adverse events (NIAE) and in full compliance with data standards issued by the Clinical Data Interchange Standards Consortium. METHODS: The validation of the VACC-Tool 2.0 (beta-version) was conducted in the context of a unique quality management program for children with suspected NIAE in collaboration with the Robert Koch Institute in Berlin, Germany. The VACC-Tool was used for instant case classification and for longitudinal follow-up throughout the course of hospitalization. Results were compared to International Classification of Diseases , Tenth Revision (ICD-10) codes assigned in the emergency department (ED). RESULTS: From 07/2013 to 10/2014, a total of 34,368 patients were seen in the ED, and 5243 patients were hospitalized; 243 of these were admitted for suspected NIAE (mean age: 8.5 years), thus participating in the quality management program. Using the VACC-Tool in the ED, 209 cases were classified successfully, 69 \\% of which had been missed or miscoded in the ED reports. Longitudinal follow-up with the VACC-Tool identified additional NIAE. CONCLUSION: Mobile applications are taking data standards to the point of care, enabling clinicians to ascertain potential adverse events in the ED setting and during inpatient follow-up. Compliance with Clinical Data Interchange Standards Consortium (CDISC) data standards facilitates data interoperability according to regulatory requirements.}, language = {en} } @inproceedings{ShaoConrad2016, author = {Shao, Borong and Conrad, Tim}, title = {Epithelial Mesenchymal Transition Regulatory Network-based Feature Selection in Lung Cancer Prognosis Prediction}, volume = {9656}, booktitle = {Lecture Notes in Computer Science (LNCS)}, doi = {10.1007/978-3-319-31744-1_13}, pages = {1235 -- 146}, year = {2016}, abstract = {Feature selection technique is often applied in identifying cancer prognosis biomarkers. However, many feature selection methods are prone to over-fitting or poor biological interpretation when applied on biological high-dimensional data. Network-based feature selection and data integration approaches are proposed to identify more robust biomarkers. We conducted experiments to investigate the advantages of the two approaches using epithelial mesenchymal transition regulatory network, which is demonstrated as highly relevant to cancer prognosis. We obtained data from The Cancer Genome Atlas. Prognosis prediction was made using Support Vector Machine. Under our experimental settings, the results showed that network-based features gave significantly more accurate predictions than individual molecular features, and features selected from integrated data (RNA-Seq and micro-RNA data) gave significantly more accurate predictions than features selected from single source data (RNA-Seq data). Our study indicated that biological network-based feature transformation and data integration are two useful approaches to identify robust cancer biomarkers.}, language = {en} } @article{TiefHoppeSeeberetal.2016, author = {Tief, Franziska and Hoppe, Christian and Seeber, L. and Obermeier, Patrick and Chen, X. and Karsch, K. and Muehlhans, S. and Adamou, E. and Conrad, Tim and Schweiger, Brunhilde and Adam, T. and Rath, Barbara}, title = {An inception cohort study assessing the role of bacterial co-infections in children with influenza and ILI and a clinical decision model for stringent antibiotic use}, volume = {21}, journal = {Antiviral Therapy}, doi = {10.3851/IMP3034}, pages = {413 -- 424}, year = {2016}, abstract = {BACKGROUND: Influenza-like illness (ILI) is a common reason for paediatric consultations. Viral causes predominate, but antibiotics are used frequently. With regard to influenza, pneumococcal coinfections are considered major contributors to morbidity/mortality. METHODS: In the context of a perennial quality management (QM) programme at the Charit{\'e} Departments of Paediatrics and Microbiology in collaboration with the Robert Koch Institute, children aged 0-18 years presenting with signs and symptoms of ILI were followed from the time of initial presentation until hospital discharge (Charit{\'e} Influenza-Like Disease = ChILD Cohort). An independent QM team performed highly standardized clinical assessments using a disease severity score based on World Health Organization criteria for uncomplicated and complicated/progressive disease. Nasopharyngeal and pharyngeal samples were collected for viral reverse transcription polymerase chain reaction and bacterial culture/sensitivity and MaldiTOF analyses. The term 'detection' was used to denote any evidence of viral or bacterial pathogens in the (naso)pharyngeal cavity. With the ChILD Cohort data collected, a standard operating procedure (SOP) was created as a model system to reduce the inappropriate use of antibiotics in children with ILI. Monte Carlo simulations were performed to assess cost-effectiveness. RESULTS: Among 2,569 ChILD Cohort patients enrolled from 12/2010 to 04/2013 (55\\% male, mean age 3.2 years, range 0-18, 19\\% {\ensuremath{>}}5 years), 411 patients showed laboratory-confirmed influenza, with bacterial co-detection in 35\\%. Influenza and pneumococcus were detected simultaneously in 12/2,569 patients, with disease severity clearly below average. Pneumococcal vaccination rates were close to 90\\%. Nonetheless, every fifth patient was already on antibiotics upon presentation; new antibiotic prescriptions were issued in an additional 20\\%. Simulation of the model SOP in the same dataset revealed that the proposed decision model could have reduced the inappropriate use of antibiotics significantly (P{\ensuremath{<}}0.01) with an incremental cost-effectiveness ratio of -99.55?. CONCLUSIONS: Physicians should be made aware that in times of pneumococcal vaccination the prevalence and severity of influenza infections complicated by pneumococci may decline. Microbiological testing in combination with standardized disease severity assessments and review of vaccination records could be cost-effective, as well as promoting stringent use of antibiotics and a personalized approach to managing children with ILI.}, language = {en} } @article{ObermeierMuehlhansHoppeetal.2016, author = {Obermeier, Patrick and Muehlhans, S. and Hoppe, Christian and Karsch, K. and Tief, Franziska and Seeber, L. and Chen, X. and Conrad, Tim and Boettcher, Sindy and Diedrich, S. and Rath, Barbara}, title = {Enabling Precision Medicine With Digital Case Classification at the Point-of-Care}, volume = {4}, journal = {EBioMedicine}, doi = {10.1016/j.ebiom.2016.01.008}, pages = {191 -- 196}, year = {2016}, abstract = {Infectious and inflammatory diseases of the central nervous system are difficult to identify early. Case definitions for aseptic meningitis, encephalitis, myelitis, and acute disseminated encephalomyelitis (ADEM) are available, but rarely put to use. The VACC-Tool (Vienna Vaccine Safety Initiative Automated Case Classification-Tool) is a mobile application enabling immediate case ascertainment based on consensus criteria at the point-of-care. The VACC-Tool was validated in a quality management program in collaboration with the Robert-Koch-Institute. Results were compared to ICD-10 coding and retrospective analysis of electronic health records using the same case criteria. Of 68,921 patients attending the emergency room in 10/2010-06/2013, 11,575 were hospitalized, with 521 eligible patients (mean age: 7.6 years) entering the quality management program. Using the VACC-Tool at the point-of-care, 180/521 cases were classified successfully and 194/521 ruled out with certainty. Of the 180 confirmed cases, 116 had been missed by ICD-10 coding, 38 misclassified. By retrospective application of the same case criteria, 33 cases were missed. Encephalitis and ADEM cases were most likely missed or misclassified. The VACC-Tool enables physicians to ask the right questions at the right time, thereby classifying cases consistently and accurately, facilitating translational research. Future applications will alert physicians when additional diagnostic procedures are required.}, language = {en} } @article{ConradBrucknerKayser2013, author = {Conrad, Tim and Bruckner, Sharon and Kayser, Bastian}, title = {Finding Modules in Networks with Non-modular Regions}, volume = {7933}, journal = {Lecture Notes in Computer Science (Proceedings of SEA 2013)}, doi = {10.1007/978-3-642-38527-8_18}, pages = {188 -- 199}, year = {2013}, abstract = {Most network clustering methods share the assumption that the network can be completely decomposed into modules, that is, every node belongs to (usually exactly one) module. Forcing this constraint can lead to misidentification of modules where none exist, while the true modules are drowned out in the noise, as has been observed e.g. for protein interaction networks. We thus propose a clustering model where networks contain both a modular region consisting of nodes that can be partitioned into modules, and a transition region containing nodes that lie between or outside modules. We propose two scores based on spectral properties to determine how well a network fits this model. We then evaluate three (partially adapted) clustering algorithms from the literature on random networks that fit our model, based on the scores and comparison to the ground truth. This allows to pinpoint the types of networks for which the different algorithms perform well.}, language = {en} } @article{ConradRathTiefetal.2013, author = {Conrad, Tim and Rath, Barbara and Tief, Franziska and Karsch, K. and Muehlhans, S. and Obermeier, Patrick and Adamou, E. and Chen, X. and Seeber, L. and Peiser, Ch. and Hoppe, Christian and von Kleist, Max and Schweiger, Brunhilde}, title = {Towards a personalized approach to managing of influenza infections in infants and children - food for thought and a note on oseltamivir}, volume = {13}, journal = {Infectious Disorders - Drug Targets}, number = {1}, pages = {25 -- 33}, year = {2013}, language = {en} } @article{ConradLeichtleNuofferetal.2012, author = {Conrad, Tim and Leichtle, Alexander Benedikt and Nuoffer, Jean-Marc and Ceglarek, Uta and Kase, Julia and Witzigmann, Helmut and Thiery, Joachim and Fiedler, Georg Martin}, title = {Serum amino acid profiles and their alterations in colorectal cancer}, journal = {Metabolomics}, doi = {10.1007/s11306-011-0357-5}, year = {2012}, abstract = {Mass spectrometry-based serum metabolic profiling is a promising tool to analyse complex cancer associated metabolic alterations, which may broaden our pathophysiological understanding of the disease and may function as a source of new cancer-associated biomarkers. Highly standardized serum samples of patients suffering from colon cancer (n = 59) and controls (n = 58) were collected at the University Hospital Leipzig. We based our investigations on amino acid screening profiles using electrospray tandem-mass spectrometry. Metabolic profiles were evaluated using the Analyst 1.4.2 software. General, comparative and equivalence statistics were performed by R 2.12.2. 11 out of 26 serum amino acid concentrations were significantly different between colorectal cancer patients and healthy controls. We found a model including CEA, glycine, and tyrosine as best discriminating and superior to CEA alone with an AUROC of 0.878 (95\\% CI 0.815?0.941). Our serum metabolic profiling in colon cancer revealed multiple significant disease-associated alterations in the amino acid profile with promising diagnostic power. Further large-scale studies are necessary to elucidate the potential of our model also to discriminate between cancer and potential differential diagnoses. In conclusion, serum glycine and tyrosine in combination with CEA are superior to CEA for the discrimination between colorectal cancer patients and controls.}, language = {en} } @article{GuptaConradSpoetteretal.2012, author = {Gupta, Pooja and Conrad, Tim and Sp{\"o}tter, Andreas and Reinsch, Norbert and Bienefeld, Kaspar}, title = {Simulating a base population in honey bee for molecular genetic studies}, journal = {Genetics Selection Evolution}, doi = {10.1186/1297-9686-44-14}, year = {2012}, abstract = {Over the past years, reports have indicated that honey bee populations are declining and that infestation by an ecto-parasitic mite (Varroa destructor) is one of the main causes. Selective breeding of resistant bees can help to prevent losses due to the parasite, but it requires that a robust breeding program and genetic evaluation are implemented. Genomic selection has emerged as an important tool in animal breeding programs and simulation studies have shown that it yields more accurate breeding values estimates, higher genetic gain and low rates of inbreeding. Since genomic selection relies on marker data, simulations conducted on a genomic dataset are a pre-requisite before selection can be implemented. Although genomic datasets have been simulated in other species undergoing genetic evaluation, simulation of a genomic dataset specific to the honey bee is required since this species has distinct genetic and reproductive biology characteristics. Our software program was aimed at constructing a base population by simulating a random mating honey bee population. A forward-time population simulation approach was applied since it allows modeling of genetic characteristics and reproductive behavior specific to the honey bee.  Results: Our software program yielded a genomic dataset for a base population in linkage disequilibrium. In addition, information was obtained on (1) the position of markers on each chromosome, (2) allele frequency, (3) ?2 statistics for Hardy- Weinberg equilibrium, (4) a sorted list of markers with a minor allele frequency less than or equal to the input value, (5) average r2 values of linkage disequilibrium between all simulated marker loci pair for all generations and (6) average r2 value of linkage disequilibrium in the last generation for selected markers with the highest minor allele frequency. Conclusion: We developed a software program that takes into account the genetic and reproductive biology characteristics specific to the honey bee and that can be used to constitute a genomic dataset compatible with the simulation studies necessary to optimize breeding programs. The source code together with an instruction file is freely accessible at http://msproteomics.org/Research/Misc/honeybeepopulationsimulator.html}, language = {en} } @article{Conrad2004, author = {Conrad, Tim}, title = {New Appraches for Visualizing and Analyzing Metabolic Pathways}, journal = {Proceedings of the Second Australian Undergraduate Students? Computing Conference}, year = {2004}, abstract = {Visualizing of metabolic pathways (or networks) has been done by many differentapproaches. In this work, we implemented and tested existing graph layout algorithms, and present a new approach to lay-out medium size metabolic pathways (500-20,000 vertices) by implementing and combining three well known graph lay-out algorithms (high dimension embedding, spring-embedder preprocessing, spring-embedder), through 3D space density analysis facilitated by the Octree technique. For the analysis of the results of metabolic pathways simulations we present two new techniques: rstly, a powerful technique to visualize pathways simulation data was created to unveil and understand concentration ows through metabolic pathways. This was achieved by mapping the color encoded concentration value of every substance from each time step of the simulation to its graphical representation in the layout. By combining all resulting images (from each time step) and displaying them as a movie, many characteristics such as subnetworks, alternative routes through the network, and differences between a modied pathway and its unmodied version can be revealed. Secondly, a new method to detect co-regulated substances in metabolic pathways and to recognize differences between two versions of a pathway, was established. To do this, we transformed the simulation data into a row-based representation, color-coded these rows, and reordered them with respect to similarity by using a Genetic Algorithm variant. From the arising discrete 2-dimensional matrix consisting of concentration values, a continuous 2-dimensional fourier row function was computed. This function can be used to measure properties, such as similarities in a pathway between time steps, or substances, or to detect and evaluate differences between modied versions of the same pathway.}, language = {en} } @phdthesis{Conrad2004, author = {Conrad, Tim}, title = {Metabolic Pathways}, year = {2004}, language = {en} } @article{VegaSchuetteConrad2016, author = {Vega, Iliusi and Sch{\"u}tte, Christof and Conrad, Tim}, title = {Finding metastable states in real-world time series with recurrence networks}, volume = {445}, journal = {Physica A: Statistical Mechanics and its Applications}, doi = {10.1016/j.physa.2015.10.041}, pages = {1 -- 17}, year = {2016}, abstract = {In the framework of time series analysis with recurrence networks, we introduce a self-adaptive method that determines the elusive recurrence threshold and identifies metastable states in complex real-world time series. As initial step, we introduce a way to set the embedding parameters used to reconstruct the state space from the time series. We set them as the ones giving the maximum Shannon entropy of the diagonal line length distribution for the first simultaneous minima of recurrence rate and Shannon entropy. To identify metastable states, as well as the transitions between them, we use a soft partitioning algorithm for module finding which is specifically developed for the case in which a system shows metastability. We illustrate our method with a complex time series example. Finally, we show the robustness of our method for identifying metastable states. Our results suggest that our method is robust for identifying metastable states in complex time series, even when introducing considerable levels of noise and missing data points.}, language = {en} } @article{ConradGenzelCvetkovicetal.2017, author = {Conrad, Tim and Genzel, Martin and Cvetkovic, Nada and Wulkow, Niklas and Leichtle, Alexander Benedikt and Vybiral, Jan and Kytyniok, Gitta and Sch{\"u}tte, Christof}, title = {Sparse Proteomics Analysis - a compressed sensing-based approach for feature selection and classification of high-dimensional proteomics mass spectrometry data}, volume = {18}, journal = {BMC Bioinfomatics}, number = {160}, doi = {10.1186/s12859-017-1565-4}, year = {2017}, abstract = {Background: High-throughput proteomics techniques, such as mass spectrometry (MS)-based approaches, produce very high-dimensional data-sets. In a clinical setting one is often interested in how mass spectra differ between patients of different classes, for example spectra from healthy patients vs. spectra from patients having a particular disease. Machine learning algorithms are needed to (a) identify these discriminating features and (b) classify unknown spectra based on this feature set. Since the acquired data is usually noisy, the algorithms should be robust against noise and outliers, while the identified feature set should be as small as possible. Results: We present a new algorithm, Sparse Proteomics Analysis (SPA),based on thet heory of compressed sensing that allows us to identify a minimal discriminating set of features from mass spectrometry data-sets. We show (1) how our method performs on artificial and real-world data-sets, (2) that its performance is competitive with standard (and widely used) algorithms for analyzing proteomics data, and (3) that it is robust against random and systematic noise. We further demonstrate the applicability of our algorithm to two previously published clinical data-sets.}, language = {en} } @misc{VegaSchuetteConrad2014, author = {Vega, Iliusi and Sch{\"u}tte, Christof and Conrad, Tim}, title = {SAIMeR: Self-adapted method for the identification of metastable states in real-world time series}, issn = {1438-0064}, url = {http://nbn-resolving.de/urn:nbn:de:0297-zib-50130}, year = {2014}, abstract = {In the framework of time series analysis with recurrence networks, we introduce SAIMeR, a heuristic self-adapted method that determines the elusive recurrence threshold and identifies metastable states in complex time series. To identify metastable states as well as the transitions between them, we use graph theory concepts and a fuzzy partitioning clustering algorithm. We illustrate SAIMeR by applying it to three real-world time series and show that it is able to identify metastable states in real-world data with noise and missing data points. Finally, we suggest a way to choose the embedding parameters used to construct the state space in which this method is performed, based on the analysis of how the values of these parameters affect two recurrence quantitative measurements: recurrence rate and entropy.}, language = {en} } @misc{SchuetteConrad2014, author = {Sch{\"u}tte, Christof and Conrad, Tim}, title = {Showcase 3: Information-based medicine}, volume = {1}, journal = {MATHEON-Mathematics for Key Technologies}, editor = {Deuflhard, Peter and Gr{\"o}tschel, Martin and H{\"o}mberg, Dietmar and Horst, Ulrich and Kramer, J{\"u}rg and Mehrmann, Volker and Polthier, Konrad and Schmidt, Frank and Skutella, Martin and Sprekels, J{\"u}rgen}, publisher = {European Mathematical Society}, pages = {66 -- 67}, year = {2014}, language = {en} } @article{MelnykMontavonKlusetal.2020, author = {Melnyk, Kateryna and Montavon, Gr{\`e}goire and Klus, Stefan and Conrad, Tim}, title = {Graph Kernel Koopman Embedding for Human Microbiome Analysis}, volume = {5}, journal = {Applied Network Science}, number = {96}, doi = {10.1007/s41109-020-00339-2}, year = {2020}, abstract = {More and more diseases have been found to be strongly correlated with disturbances in the microbiome constitution, e.g., obesity, diabetes, or some cancer types. Thanks to modern high-throughput omics technologies, it becomes possible to directly analyze human microbiome and its influence on the health status. Microbial communities are monitored over long periods of time and the associations between their members are explored. These relationships can be described by a time-evolving graph. In order to understand responses of the microbial community members to a distinct range of perturbations such as antibiotics exposure or diseases and general dynamical properties, the time-evolving graph of the human microbial communities has to be analyzed. This becomes especially challenging due to dozens of complex interactions among microbes and metastable dynamics. The key to solving this problem is the representation of the time-evolving graphs as fixed-length feature vectors preserving the original dynamics. We propose a method for learning the embedding of the time-evolving graph that is based on the spectral analysis of transfer operators and graph kernels. We demonstrate that our method can capture temporary changes in the time-evolving graph on both synthetic data and real-world data. Our experiments demonstrate the efficacy of the method. Furthermore, we show that our method can be applied to human microbiome data to study dynamic processes.}, language = {en} } @article{IravaniConrad2023, author = {Iravani, Sahar and Conrad, Tim}, title = {An Interpretable Deep Learning Approach for Biomarker Detection in LC-MS Proteomics Data}, volume = {20}, journal = {IEEE/ACM Transactions on Computational Biology and Bioinformatics}, number = {1}, doi = {10.1109/tcbb.2022.3141656}, pages = {151 -- 161}, year = {2023}, abstract = {Analyzing mass spectrometry-based proteomics data with deep learning (DL) approaches poses several challenges due to the high dimensionality, low sample size, and high level of noise. Additionally, DL-based workflows are often hindered to be integrated into medical settings due to the lack of interpretable explanation. We present DLearnMS, a DL biomarker detection framework, to address these challenges on proteomics instances of liquid chromatography-mass spectrometry (LC-MS) - a well-established tool for quantifying complex protein mixtures. Our DLearnMS framework learns the clinical state of LC-MS data instances using convolutional neural networks. Based on the trained neural networks, we show how biomarkers can be identified using layer-wise relevance propagation. This enables detecting discriminating regions of the data and the design of more robust networks. One of the main advantages over other established methods is that no explicit preprocessing step is needed in our DLearnMS framework. Our evaluation shows that DLearnMS outperforms conventional LC-MS biomarker detection approaches in identifying fewer false positive peaks while maintaining a comparable amount of true positives peaks.}, language = {en} } @article{RamsConrad2022, author = {Rams, Mona and Conrad, Tim}, title = {Dictionary learning allows model-free pseudotime estimation of transcriptomics data}, volume = {23}, journal = {BMC Genomics}, publisher = {BioMed Central}, doi = {10.1186/s12864-021-08276-9}, year = {2022}, language = {en} } @article{WeimannConrad2021, author = {Weimann, K. and Conrad, Tim}, title = {Transfer Learning for ECG Classification}, volume = {11}, journal = {Scientific Reports}, doi = {10.1038/s41598-021-84374-8}, year = {2021}, abstract = {Remote monitoring devices, which can be worn or implanted, have enabled a more effective healthcare for patients with periodic heart arrhythmia due to their ability to constantly monitor heart activity. However, these devices record considerable amounts of electrocardiogram (ECG) data that needs to be interpreted by physicians. Therefore, there is a growing need to develop reliable methods for automatic ECG interpretation to assist the physicians. Here, we use deep convolutional neural networks (CNN) to classify raw ECG recordings. However, training CNNs for ECG classification often requires a large number of annotated samples, which are expensive to acquire. In this work, we tackle this problem by using transfer learning. First, we pretrain CNNs on the largest public data set of continuous raw ECG signals. Next, we finetune the networks on a small data set for classification of Atrial Fibrillation, which is the most common heart arrhythmia. We show that pretraining improves the performance of CNNs on the target task by up to 6.57\%, effectively reducing the number of annotations required to achieve the same performance as CNNs that are not pretrained. We investigate both supervised as well as unsupervised pretraining approaches, which we believe will increase in relevance, since they do not rely on the expensive ECG annotations. The code is available on GitHub at https://github.com/kweimann/ecg-transfer-learning.}, language = {en} } @article{LeDucConrad2020, author = {Le Duc, Huy and Conrad, Tim}, title = {A light-weight and highly flexible software system for analyzing large bio-medical datasets}, journal = {Future Generation Computer Systems}, year = {2020}, language = {en} } @article{JudsSchmidtWelleretal.2020, author = {Juds, Carmen and Schmidt, Johannes and Weller, Michael and Lange, Thorid and Conrad, Tim and Boerner, Hans}, title = {Combining Phage Display and Next-generation Sequencing for Materials Sciences: A Case Study on Probing Polypropylene Surfaces}, volume = {142}, journal = {Journal of the American Chemical Society}, number = {24}, doi = {10.1021/jacs.0c03482}, pages = {10624 -- 10628}, year = {2020}, abstract = {Phage display biopanning with Illumina next-generation sequencing (NGS) is applied to reveal insights into peptide-based adhesion domains for polypropylene (PP). One biopanning round followed by NGS selects robust PP-binding peptides that are not evident by Sanger sequencing. NGS provides a significant statistical base that enables motif analysis, statistics on positional residue depletion/enrichment, and data analysis to suppress false-positive sequences from amplification bias. The selected sequences are employed as water-based primers for PP?metal adhesion to condition PP surfaces and increase adhesive strength by 100\\% relative to nonprimed PP.}, language = {en} } @article{CvetkovicConradLie2021, author = {Cvetkovic, Nada and Conrad, Tim and Lie, Han Cheng}, title = {A Convergent Discretisation Method for Transition Path Theory for Diffusion Processes}, volume = {19}, journal = {Multiscale Modeling \& Simulation}, number = {1}, publisher = {Society for Industrial and Applied Mathematics}, doi = {10.1137/20M1329354}, pages = {242 -- 266}, year = {2021}, language = {en} } @article{ZhangKlusConradetal.2019, author = {Zhang, Wei and Klus, Stefan and Conrad, Tim and Sch{\"u}tte, Christof}, title = {Learning chemical reaction networks from trajectory data}, volume = {18}, journal = {SIAM Journal on Applied Dynamical Systems (SIADS)}, number = {4}, arxiv = {http://arxiv.org/abs/1902.04920}, doi = {10.1137/19M1265880}, pages = {2000 -- 2046}, year = {2019}, abstract = {We develop a data-driven method to learn chemical reaction networks from trajectory data. Modeling the reaction system as a continuous-time Markov chain and assuming the system is fully observed,our method learns the propensity functions of the system with predetermined basis functions by maximizing the likelihood function of the trajectory data under l^1 sparse regularization. We demonstrate our method with numerical examples using synthetic data and carry out an asymptotic analysis of the proposed learning procedure in the infinite-data limit.}, language = {en} } @inproceedings{IravaniConrad2019, author = {Iravani, Sahar and Conrad, Tim}, title = {Deep Learning for Proteomics Data for Feature Selection and Classification}, volume = {11713}, booktitle = {Machine Learning and Knowledge Extraction. CD-MAKE 2019}, editor = {Holzinger, A. and Kieseberg, P. and Tjoa, A. and Weippl, E.}, publisher = {Springer, Cham}, doi = {10.1007/978-3-030-29726-8_19}, year = {2019}, language = {en} }