Zusammenfassung
Schmerz hat eine komplexe Pathophysiologie, die sich in komplexen und heterogenen klinischen Phänotypen ausdrückt. Dies macht die Erforschung von Schmerz und seiner Behandlung zu einem potenziell datenintensiven Thema, bei dessen Bearbeitung große Mengen komplexer Daten aufgenommen werden. Typische Quellen solcher Daten sind Untersuchungen mit funktioneller Magnetresonanztomographie, komplexen quantitativ-sensorischen Tests, DNA-Sequenzierung, insbesondere dem sog. „next generation sequencing“, oder funktionell-genomischen Forschungsansätzen wie zum Beispiel solchen, die auf die Entdeckung oder Repositionierung von bekannten Arzneimitteln als neue Analgetika gerichtet sind. Die Extraktion von Informationen aus solchen Big Data erfordert datenwissenschaftliche Methoden, die der Informatik mehr als der Statistik zuzuordnen sind. Derzeit richtet sich ein besonderes Interesse auf das maschinelle Lernen, welches Methoden zur Detektion interessanter, insbesondere biologisch aussagekräftiger Strukturen in hochdimensionalen Daten bereitstellt, um sog. Klassifikatoren zu erstellen, die klinische Phänotypen z. B. aus klinischen oder genetischen Merkmalen vorhersagen. Darüber hinaus können diese Methoden zur Wissensentdeckung in großen aus biomedizinischen Datenbanken ausgelesenen Datensätzen verwendet werden, um Hypothesen zu generieren und das derzeitige Wissen über Schmerz zur Entwicklung neuer Analgetika zu nutzen. Dies ermöglicht, in der Schmerzforschung sogenannte DIKW-Ansätze (Daten – Information – Wissen [„knowledge“] – Weisheit) zu verfolgen. In diesem Artikel wird anhand aktueller Beispiele aus der Schmerzforschung ein Überblick über die aktuellen datenwissenschaftlichen Methoden in diesem Forschungsbereich vermittelt.
Abstract
Pain has a complex pathophysiology that is expressed in multifaceted and heterogeneous clinical phenotypes. This makes research on pain and its treatment a potentially data-rich field as large amounts of complex data are generated. Typical sources of such data are investigations with functional magnetic resonance imaging, complex quantitative sensory testing, next-generation DNA sequencing and functional genomic research approaches, such as those aimed at analgesic drug discovery or repositioning of drugs known from other indications as new analgesics. Extracting information from these big data requires complex data scientific-based methods belonging more to computer science than to statistics. A particular interest is currently focused on machine learning, the methods of which are used for the detection of interesting and biologically meaningful structures in high-dimensional data. Subsequently, classifiers can be created that predict clinical phenotypes from, e.g. clinical or genetic features acquired from subjects. In addition, knowledge discovery in big data accessible in electronic knowledge bases, can be used to generate hypotheses and to exploit the accumulated knowledge about pain for the discovery of new analgesic drugs. This enables so-called data—information—knowledge—wisdom (DIKW) approaches to be followed in pain research. This article highlights current examples from pain research to provide an overview about contemporary data scientific methods used in this field of research.
Literatur
Basbaum AI, Bautista DM, Scherrer G et al (2009) Cellular and molecular mechanisms of pain. Cell 139:267–284
Breiman L (2001) Random forests. Mach Learn 45:5–32
Breimann L, Friedman JH, Olshen RA et al (1993) Classification and regression trees. Chapman and Hall, Boca Raton
Chollet F, Allaire JJ (2018) Deep learning with R. Manning Publications Co, Shelter Island
Cover T, Hart P (1967) Nearest neighbor pattern classification. IEEE Trans Inf Theory 13:21–27
Derry S, Gill D, Phillips T et al (2012) Milnacipran for neuropathic pain and fibromyalgia in adults. Cochrane Database Syst Rev. https://doi.org/10.1002/14651858.CD008244.pub2
Ho TK (1995) Random decision forests. In: IEEE Computer Society (Hrsg) Proceedings of the third international conference on document analysis and recognition, Bd. 1. IEEE Computer Society, Volume, S 278
Hu L, Iannetti GD (2016) Painful issues in pain prediction. Trends Neurosci 39:212–220
Julius D, Basbaum AI (2001) Molecular mechanisms of nociception. Nature 413:203–210
Kohonen T (1982) Self-organized formation of topologically correct feature maps. Biol Cybernet 43:59–69
Kringel D, Geisslinger G, Resch E et al (2018) Machine-learned analysis of the association of next-generation sequencing-based human TRPV1 and TRPA1 genotypes with the sensitivity to heat stimuli and topically applied capsaicin. Pain 159:1366–1381
Kringel D, Lötsch J (2015) Pain research funding by the European Union Seventh Framework Programme. Eur J Pain 19:595–600
Lötsch J, Geisslinger G (2010) Bedside-to-bench pharmacology: a complementary concept to translational pharmacology. Clin Pharmacol Ther 87:647–649
Lötsch J, Geisslinger G (2011) Pharmacogenetics of new analgesics. Br J Pharmacol 163:447–460
Lötsch J, Geisslinger G, Heinemann S, Lerch F, Oertel BG, Ultsch A (2017) QST response patterns to capsaicin- and UV-B-induced local skin hypersensitization in healthy subjects: a machine-learned analysis. Pain 159(1):11–24. https://doi.org/10.1097/j.pain.0000000000001008
Lötsch J, Oertel BG, Ultsch A (2014) Human models of pain for the prediction of clinical analgesia. Pain. https://doi.org/10.1016/j.pain.2014.07.003
Lotsch J, Ultsch A (2017) Machine learning in pain research. Pain 159:623–630
Mayer EA, Gupta A, Kilpatrick LA et al (2015) Imaging brain mechanisms in chronic visceral pain. Pain 156(Suppl 1):S50–S63
Mogil JS (2009) Animal models of pain: progress and challenges. Nat Rev Neurosci 10:283–294
Moore RA, Derry S, Aldington D et al (2012) Amitriptyline for neuropathic pain and fibromyalgia in adults. Cochrane Database Syst Rev. https://doi.org/10.1002/14651858.CD008242.pub3
Moore RA, Straube S, Wiffen PJ et al (2009) Pregabalin for acute and chronic pain in adults. Cochrane Database Syst Rev. https://doi.org/10.1002/14651858.CD007076.pub2
Oertel BG, Lötsch J (2013) Clinical pharmacology of analgesics assessed with human experimental pain models: bridging basic and clinical research. Br J Pharmacol 168:534–553
Oertel BG, Preibisch C, Wallenhorst T et al (2008) Differential opioid action on sensory and affective cerebral pain processing. Clin Pharmacol Ther 83:577–588
President’s Information Technology Advisory C (2005) Report to the president: computational science: ensuring America’s competitiveness
Rice AS, Cimino-Brown D, Eisenach JC et al (2008) Animal models and the prediction of efficacy in clinical trials of analgesic drugs: a critical appraisal and call for uniform reporting standards. Pain 139:243–247
Rolke R, Baron R, Maier C et al (2006) Quantitative sensory testing in the German Research Network on Neuropathic Pain (DFNS): standardized protocol and reference values. Pain 123:231–243
Schapire RE, Freund Y (1999) A short introduction to boosting. J Japanese Soc Artif Intell 14:771–780
Thrun MC (2017) A system for projection based clustering through self-organization and swarm intelligence. PhD thesis, Philipps-University, Marburg. Springer, Heidelberg
Ultsch A (2003) The U‑matrix as visualization for projections of high-dimensional data. In: Locarek-Junge H (Hrsg) Proc. 11th IFCS Biennial Conference
Ultsch A, Herrmann L (2010) Self organized swarms for cluster preserving projections of high-dimensional data. In: ECEASST
Ultsch A, Lötsch J (2017) Machine-learned cluster identification in high-dimensional data. J Biomed Inform 66:95–104
Ultsch A, Sieman HP (1990) Kohonen’s self organizing feature maps for exploratory data analysis. In: INNC’90, Int. Neural Network Conference. Kluwer, Dordrecht, S 305–308
Von Hehn CA, Baron R, Woolf CJ (2012) Deconstructing the neuropathic pain phenotype to reveal neural mechanisms. Neuron 73:638–652
Walker SH, Duncan DB (1967) Estimation of the probability of an event as a function of several independent variables. Biometrika 54:167–179
Walter C, Oertel BG, Felden L et al (2016) Brain mapping-based model of delta(9)-tetrahydrocannabinol effects on connectivity in the pain matrix. Neuropsychopharmacology 41:1659–1669
Wickham H, Grolemund G (2017) R for data science: import, tidy, transform, visualize, and model data. O’Reilly Media, Sebastopol, CA 95472, USA
Förderung
Landesoffensive zur Entwicklung wissenschaftlich-ökonomischer Exzellenz (LOEWE), LOEWE-Zentrum für Translationale Medizin und Pharmakologie (GG, JL)
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Interessenkonflikt
J. Lötsch, G. Geisslinger und C. Walter geben an, dass kein Interessenkonflikt besteht.
Für diesen Beitrag wurden von den Autoren keine Studien an Menschen oder Tieren durchgeführt. Für die aufgeführten Studien gelten die jeweils dort angegebenen ethischen Richtlinien.
Rights and permissions
About this article
Cite this article
Lötsch, J., Geisslinger, G. & Walter, C. Wissensgenerierung aus komplexen Datensätzen in der humanexperimentellen Schmerzforschung. Schmerz 33, 502–513 (2019). https://doi.org/10.1007/s00482-019-00412-5
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00482-019-00412-5
Schlüsselwörter
- Humanexperimentelle Schmerzforschung
- „Data science“
- Klinische Pharmakologie
- Maschinelles Lernen
- Pharmakometrie