Application of Clinical Bioinformatics

395 Pages·2016·6.905 MB·English

by  Xiangdong Wang, Christian Baumgartner, Denis C. Shields, Hong-Wen Deng, Jacques S Beckmann (eds.)

Translational Bioinformatics 11
Series Editor: Xiangdong Wang, MD, Ph.D., Prof
Xiangdong Wang
Christian Baumgartner
Denis C. Shields
Hong-Wen Deng
Jacques S. Beckmann Editors
Application
of Clinical
Bioinformatics
Translational Bioinformatics
Volume 11
Series editor
Xiangdong Wang, MD, Ph.D.
Professor of Medicine, Zhongshan Hospital, Fudan University Medical School,
China
Director of Shanghai Institute of Clinical Bioinformatics, (www.fuccb.org)
AimsandScope
The Book Series in Translational Bioinformatics is a powerful and integrative resource for
understanding and translating discoveries and advances of genomic, transcriptomic, proteomic
andbioinformatictechnologiesintothestudyofhumandiseases.TheSeriesrepresentsleading
global opinions on the translation of bioinformatics sciences into both the clinical setting and
descriptions to medical informatics. It presents the critical evidence to further understand the
molecular mechanisms underlying organ or cell dysfunctions in human diseases, the results of
genomic,transcriptomic,proteomicandbioinformaticstudiesfromhumantissuesdedicatedtothe
discoveryandvalidationofdiagnosticandprognosticdiseasebiomarkers,essentialinformationon
the identification and validation of novel drug targets and the application of tissue genomics,
transcriptomics,proteomicsandbioinformaticsindrugefficacyandtoxicityinclinicalresearch.
The Book Series in Translational Bioinformatics focuses on outstanding articles/chapters
presenting significant recent works in genomic, transcriptomic, proteomic and bioinformatic
profiles related to human organ or cell dysfunctions and clinical findings. The Series includes
bioinformatics-driven molecular and cellular disease mechanisms, the understanding of human
diseasesandtheimprovementofpatientprognoses.Additionally,itprovidespracticalanduseful
studyinsightsintoandprotocolsofdesignandmethodology.
SeriesDescription
Translational bioinformatics is defined as the development of storage-related, analytic, and
interpretivemethodstooptimizethetransformationofincreasinglyvoluminousbiomedicaldata,
and genomic data in particular, into proactive, predictive, preventive, and participatory health.
Translational bioinformatics includes research on the development of novel techniques for the
integrationofbiologicalandclinicaldataandtheevolutionofclinicalinformaticsmethodologyto
encompassbiologicalobservations.Theendproductoftranslationalbioinformaticsisthenewly
found knowledge from these integrative efforts that can be disseminated to a variety of stake-
holdersincludingbiomedicalscientists,clinicians,andpatients.Issuesrelatedtodatabaseman-
agement,administration,orpolicywillbecoordinatedthroughtheclinicalresearchinformatics
domain.Analytic,storage-related,andinterpretivemethodsshouldbeusedtoimprovepredictions,
earlydiagnostics,severitymonitoring,therapeuticeffects,andtheprognosisofhumandiseases.
RecentlyPublishedandForthcomingVolumes
ComputationalandStatistical AllergyBioinformatics
Epigenomics Editors:AilinTao,EyalRaz
Editor:AndrewE.Teschendorff Volume8
Volume7
PediatricBiomedicalInformatics-
TranscriptomicsandGeneRegulation ComputerApplications
Editor:JiaqianWu inPediatricResearch(Edition2)
Volume9 Editor:JohnJ.Hutton
Volume10
Moreinformationaboutthisseriesathttp://www.springer.com/series/11057
Xiangdong Wang • Christian Baumgartner
Denis C. Shields • Hong-Wen Deng
Jacques S. Beckmann
Editors
Application of Clinical
Bioinformatics
Editors
XiangdongWang ChristianBaumgartner
ZhongshanHospital,FudanUniversity InstituteofHealthCareEngineeringwith
ShanghaiInstituteofClinical EuropeanNotifiedBodyofMedicalDevices
Bioinformatics GrazUniversityofTechnology
Shanghai,China Graz,Austria
DenisC.Shields Hong-WenDeng
SchoolofMedicine CenterforBioinformaticsandGenomics,
UniversityCollegeDublin DepartmentofBiostatistics
Dublin4,Ireland andBioinformatics
TulaneUniversitySchoolofPublicHealth
JacquesS.Beckmann andTropicalMedicine
SectionofClinicalBioinformatics NewOrleans,LA,USA
SwissInstituteofBioinformatics
Switzerland
ISSN2213-2775 ISSN2213-2783 (electronic)
ISBN978-94-017-7541-0 ISBN978-94-017-7543-4 (eBook)
DOI10.1007/978-94-017-7543-4
LibraryofCongressControlNumber:2016933682
SpringerDordrechtHeidelbergNewYorkLondon
©SpringerScience+BusinessMediaDordrecht2016
Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartof
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilarmethodologynowknownorhereafterdeveloped.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publicationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexempt
fromtherelevantprotectivelawsandregulationsandthereforefreeforgeneraluse.
Thepublisher,theauthorsandtheeditorsaresafetoassumethattheadviceandinformationinthis
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained
hereinorforanyerrorsoromissionsthatmayhavebeenmade.
Printedonacid-freepaper
Springer Science+Business Media B.V. Dordrecht is part of Springer Science+Business Media
(www.springer.com)
Contents
1 TheEraofBigData:FromData-DrivenResearch
toData-DrivenClinicalCare. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
ChristianBaumgartner
2 Biostatistics,DataMiningandComputationalModeling. . . . . . . . 23
HaoHe,DongdongLin,JigangZhang,
YupingWang,andHong-WenDeng
3 GeneExpressionandProfiling. . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
YuZhou,ChaoXu,JigangZhang,andHong-WenDeng
4 TheNextGenerationSequencingandApplications
inClinicalResearch. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
JunboDuan,XiaoyingFu,JigangZhang,
Yu-PingWang,andHong-WenDeng
5 ClinicalEpigeneticsandEpigenomics. . . . . . . . . . . . . . . . . . . . . . . 115
NianDong,LinShi,ChengshuiChen,WenhuanMa,
andXiangdongWang
6 ProteomicProfiling:DataMiningandAnalyses. . . . . . . . . . . . . . . 133
LanZhang,WeiZhu,YongZeng,
JigangZhang,andHong-WenDeng
7 TargetedMetabolomics:TheNextGeneration
ofClinicalChemistry!. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
KlausM.WeinbergerandMarcBreit
8 ClinicalBioinformaticsforBiomarkerDiscoveryinTargeted
Metabolomics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
MarcBreit,ChristianBaumgartner,MichaelNetzer,
andKlausM.Weinberger
v
vi Contents
9 MetagenomicProfiling,InteractionofGenomics
withMeta-genomics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
RuifengWang,YuZhou,ShaolongCao,YupingWang,
JigangZhang,andHong-WenDeng
10 ClinicalEpigeneticsandEpigenomics. . . . . . . . . . . . . . . . . . . . . . . 269
ChuanQiu,FangtangYu,Hong-WenDeng,andHuiShen
11 IntegrativeBiologicalDatabases. . . . . . . . . . . . . . . . . . . . . . . . . . . 295
JinzengWangandHaiyunWang
12 StandardsandRegulationsfor(Bio)MedicalSoftware. . . . . . . . . . 309
J€orgSchr€ottner,RobertNeubauer,andChristianBaumgartner
13 ClinicalApplicationsandSystemsBiomedicine. . . . . . . . . . . . . . . 323
DuojiaoWu,DavidE.Sanin,andXiangdongWang
14 KeyLawandPolicyConsiderations
forClinicalBioinformaticians. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
MarkPhillips
15 ChallengesandOpportunitiesinClinicalBioinformatics. . . . . . . . 359
DenisC.Shields
16 HeterogeneityofHepatocellularCarcinoma. . . . . . . . . . . . . . . . . . 371
TingtingFang,LiFeng,andJinglinXia
Chapter 1
The Era of Big Data: From Data-Driven
Research to Data-Driven Clinical Care
ChristianBaumgartner
Abstract Whentheeraofbigdataarrivedintheearlynineteennineties,biomed-
ical research boosted new innovations, procedures and methods aiding in clinical
care and patient management. This chapter provides an introduction to the basic
conceptsandstrategiesofdata-drivenbiomedicalresearchandapplication,anarea
thatisexplainedusingtermssuchascomputationalbiomedicineorclinical/medical
bioinformatics.Afterabriefmotivationitstartswithasurveyondatasourcesand
bioanalytictechnologiesforhigh-throughputdatageneration,aselectionofexper-
imental study designs andtheir applications, proceduresandrecommendationson
how to handle data quality and privacy, followed by a discussion on basic data
warehouse concepts utilized for life science data integration, data mining and
knowledge discovery. Finally, five application examples are briefly delineated,
emphazising the benefit and power of computational methods and tools in this
field. The author trusts that this chapter will encourage the reader to handle and
interpretthehugeamountofdatausuallygeneratedinresearchprojectsorclinical
routinetoexploitminedbioinformationandmedicalknowledgeforindividualized
healthcare.
Keywords Computational biomedicine • Data integration and management •
Knowledgediscovery•Datamining•Clinicalapplications
1.1 Introduction
In the past two decades, the new era of “big data” in experimental and clinical
biomedicine has arrived and grown as a direct consequence of the availability of
largereservoirsofdata.Datacollectionindigitalformwasalreadyunderwaybythe
1960s, allowing for retrospective data management and analysis to be undertaken
using computers for the first time. Relational databases arose in the 1980s along
withStructuredQueryLanguages(SQL),enablingdynamic,on-demandstructural
analysisandinterpretationofdatafromcomplexresearchdesigns.The1990ssaw
C.Baumgartner,Ph.D.(*)
InstituteofHealthCareEngineeringwithEuropeanNotifiedBodyofMedicalDevices,Graz
UniversityofTechnology,Stremayrgasse16,A-8010Graz,Austria
e-mail:[email protected]
©SpringerScience+BusinessMediaDordrecht2016 1
X.Wangetal.(eds.),ApplicationofClinicalBioinformatics,
TranslationalBioinformatics11,DOI10.1007/978-94-017-7543-4_1
2 C.Baumgartner
anexplosioninthe growth ofdataassociated with theemerginguse ofnew high-
throughput,labandimagingtechnologiesinfundamentalbiomedicalresearchand
clinical application. Data warehouses were beginning to be used for storing and
integratingvarioustypesofdata,wheredifferentdatasourcesaretransformedinto
a common format and converted to a common vocabulary needed to overcome
computationalchallengesofdata-drivenresearchanddevelopment.Theneweraof
“computational biomedicine” or “clinical bioinformatics” was born as a multidis-
ciplinary approach that brought together medical, natural and computer sciences,
aiming at uncovering unknown and unexpected biomedical knowledge stored in
these data sources, which had the potential to transform our current clinical
practices (Chang 2005; Wang and Liotta 2011; Coveney et al. 2014). Research
areas such as data warehousing and information retrieval, machine learning, data
mining, and others thus arose as a response to challenges faced by the computer
science and bioinformatics community in dealing with huge amounts of data,
enabling a better quality of data-driven decision making. As data are any facts,
numbers, images or texts that can be accessed and processed by stand-alone
computers or computational networks, the patterns, associations or relationships
among available data can provide informationabout historical patternsand future
trends so that undreamt of opportunities emerge for biomedical research and
application.Thisknowledgemayhelptocreateanewwayofdealingwithclinical
care and patient management never previously possible. Clinical bioinformatics,
which resulted from the big data era, is thus a crucial element of the medical
knowledge discovery process where relevant sources of medical information and
bioinformationarecombinedandminedtoallowforindividualizedhealthcare.
1.2 The Revolution of High-Throughput and Imaging
Technologies and the Flood of Generated Data
Inthelifesciences,hugeamountofdataaregenerated,utilizingthewidespectrumof
high throughput and laboratory technologies, and modern health care imaging sys-
tems such as MRI or CT. In biomolecular research, microarray based expression
profiling and more recently next-generation sequencing (NGS) technologies have
becomethemethodologyofchoicee.g.forwholetranscriptomeexpressionprofiling,
producing a flood of data that need to be computationally processed and analysed
(Worthey2013;Soonetal.2013).ThemostwidelyusedNGSdevices,forexample,
areabletosequenceupto150basesfrombothsidesofRNAfragmentsandcreatea
maximum output of up to 1000 GB per run. Most advanced protein profiling
technologiesareimplementedwithabroadpanelofmassspectrometry-basedtech-
niques to separate, characterize and quantify analytes from complex biological
samples(ChenandPramanik2009;BrewisandBrennan2010;Woodsetal.2014).
Labsaretypicallyequippedwithdiversemassspectrometer(MS)systemsincluding
TOF-TOF, Quadrupole-TOF, FT-ICR, and LTQ-Orbitrap type analyzers. In this
field, shotgun proteomics is a widely used tool for global analysis of protein
1 TheEraofBigData 3
modifications,where,inatypicalLC-MS/MSexperiment,hundredsofthousandsof
tandem mass spectra are typically generated. Sophisticated computational tools for
MSsprectraprocessinganddatabasesearchstrategiesareusedfortheidentification
of peptide/protein modifications (Baumgartner et al. 2008; Cerqueira et al. 2010;
Sj€ostr€om et al. 2015). In metabolomics, different fundamental approaches can be
distinguished,i.e.untargetedandtargetedmetabolomicsandmetabolicfingerprinting
(BaumgartnerandGraber2007;Putrietal.2013;Nazetal.2014;Zhangetal.2015).
Usingtargetedmetabolomics,quantitiationofapreselectedsetofknownmetabolites
by determining absolute values of analyte concentrations with the use of internal
chemical standards allows for hypothesis-driven research and interpretation ofdata
basedona-prioriknowledge.Toprovideaholisticpictureofmetabolism,untargeted
metabolic profiling aims at measuring as many analytes as possible (up to several
hundreds)tocreateasnapshotofthebiochemicalprofilewithintheanalysedsample.
Theestablishedtechnologiesinmetabolomicsinclude–analoguoustoproteomics–
massspectrometrybasedapproachesandnuclearmagneticresonance(NMR)spec-
troscopy, generatingthousands totensofthousands data pointsper spectrum. Mul-
tiple processing steps are required to analyze this huge amount of spectral
information, ranging from modalities for denoising, binning, aligning spectra to
peakdetectionandhigh-levelanalysise.g.forbiomarkeridentificationandverifica-
tion(Swanetal.2013;Netzeretal.2015).
Nowadays bioimaging devices with increasing resolution are widely used in
biological and clinical laboratories, generating imaging data with hundreds of
Megabytes or Gigabytes (Eliceiri et al. 2012; Edelstein et al. 2014). Whole-slide
bioimaging, for instance, combines light microscopy techniques with electronic
scanningofslidesandisabletocollectquantitativedata,currentlyregardedasone
of the most promising avenues for diagnosis or prediction of cancer and other
diseases.TraditionalhealthcareimagingtechnologiessuchasCT,MRI,ultrasound
or SPECT and PET make it possible to assess the current status and condition of
organsortissuesandtomonitorpatientsovertimefordiagnosticevaluationorfor
controlling therapeutic interventions (Smith and Webb 2010; Mikla and Mikla
2013).Inparticular,CPU-intensiveimagereconstructionandmodelingtechniques
allow instant processing of2D signals tocreate 3D/4D image stacks of enormous
amounts of data, typically stored in DICOM file format. This DICOM standard
facilitates interoperability of medical imaging instrumentations, providing a stan-
dardized medical file format and directory structure, which enables access to the
images and patient-related information for further processing, modeling and
analysis.
1.3 Study Design and Data Privacy
Differentepidemiologicalstudydesignssuchascase-control,(longitudinal)cohort
studiesormorecomplexdesignssuchasrandomizedcontrolledtrialsareselected
inbiomedicalresearch(DawsonandTrapp2004;Porta2014).Case-controlstudies,