REVIEW
Human-based app oaches o pha macology and
ca diology: an in e disciplina y and in e sec o ial
wo kshop
Blanca Rod iguez1*, Annama ia Ca usi2, Najah Abi-Ge ges3, Rina A iga4,
Oli e B i on1, Gil Bub5, Al onso Bueno-O o io1, Rebecca A.B. Bu on5,
Valen ina Ca apella6, Louie Ca done-Noo 1, Ma hew J. Daniels4, Ma k R. Da ies7,
Sa a Du a1, And e Ghe i3, Vicen e G au6, S ephen Ha me 8, I an Koplja 9,
Pie Lambiase10, Hua Rong Lu9, Au o e Lyon1, Ana Minchole1, Anna Muszkiewicz1,
Julien Os e 6, Michelangelo Paci11, Elisa Passini1, S e ano Se e i12, Pe e Tagga 10,
Andy Tinke 8, Jean-Pie e Valen in13, And as Va o14, Mikael Wallman15, and
Xin Zhou1
1
Depa men o Compu e Science, Uni e si y o Ox o d, Ox o d, UK;
2
Medical Humani ies, Uni e si y o She ield, She ield, UK;
3
AnaBios Co po a ion, San Diego Science Cen e ,
San Diego, CA 92109, USA;
4
Radcli e Depa men o Medicine, Uni e si y o Ox o d, John Radcli e Hospi al, Ox o d, UK;
5
Depa men o Physiology, Ana omy and Gene ics,
Uni e si y o Ox o d, Ox o d, UK;
6
Depa men o Enginee ing Science, Uni e si y o Ox o d, Ox o d, UK;
7
QT-In o ma ics Limi ed, Maccles ield, UK;
8
William Ha ey Hea Cen e,
Ba s and The London School o Medicine and Den is y, Queen Ma y Uni e si y o London, Cha e house Squa e, London, UK;
9
Disco e y Sciences, Dis&De Resea ch, Janssen
Pha maceu ical NV, Bee se, Belgium;
10
Ins i u e o Ca dio ascula Science, Uni e si y College London, Ba s Hea Cen e, London, UK;
11
Depa men o Elec onics and
Communica ions Enginee ing, Tampe e Uni e si y o Technology, BioMediTech, Tampe e, Finland;
12
Depa men o Elec ical, Elec onic and In o ma ion Enginee ing, Uni e si y o
Bologna, Cesena 47521, I aly;
13
Non-Clinical De elopmen , UCB-Biopha ma, B aine L’Alleud B-1420, Belgium;
14
Uni e si y o Szeged, Szeged, Hunga y; and
15
F aunho e -Chambe s
Cen e, Go henbu g, Sweden
Recei ed 11 June 2015; accep ed a e e ision 20 Augus 2015; online publish-ahead-o -p in 29 No embe 2015
Bo h biomedical esea ch and clinical p ac ice ely on complex da ase s o he physiological and gene ic cha ac e iza ion o human hea s in
heal h and disease. Gi en he complexi y and a ie y o app oaches and eco dings, he e is now g owing ecogni ion o he need o embed
compu a ional me hods in ca dio ascula medicine and science o analysis, in eg a ion and p edic ion. This pape desc ibes a Wo kshop on
Compu a ional Ca dio ascula Science ha c ea ed an in e na ional, in e disciplina y and in e -sec o ial o um o de ine he nex s eps o a
human-based app oach o disease suppo ed by compu a ional me hodologies. The main ideas highligh ed we e (i) a shi owa ds human-
based me hodologies, spu ed by ad ances in new in silico,in i o,in i o, and ex i o echniques and he inc easing acknowledgemen o he
limi a ions o animal models. (ii) Compu a ional app oaches complemen , expand, b idge, and in eg a e in i o,in i o, and ex i o expe imen al
and clinical da a and me hods, and as such hey a e an in eg al pa o human-based me hodologies in pha macology and medicine. (iii) The
e ec i e implemen a ion o mul i- and in e disciplina y app oaches, eams, and aining combining and in eg a ing compu a ional me hods wi h
expe imen al and clinical app oaches ac oss academia, indus y, and heal hca e se ings is a p io i y. (i ) The human-based c oss-disciplina y
app oach equi es expe s in speci ic me hodologies and domains, who also ha e he capaci y o communica e and collabo a e ac oss disci-
plines and c oss-sec o en i onmen s. ( ) This new ansla ional domain o human-based ca diology and pha macology equi es new pa ne -
ships suppo ed inancially and ins i u ionally ac oss sec o s. Ins i u ional, o ganiza ional, and social ba ie s mus be iden i ied, unde s ood and
o e come in each speci ic se ing.
-----------------------------------------------------------------------------------------------------------------------------------------------------------
Keywo ds Human-based me hods †Compu a ional app oaches †Human elec ophysiology †S em-cell-de i ed
ca diomyocy es †Compu e modelling and simula ions †A hy hmias †Bioma ke s
*Co esponding au ho . Tel: +44 0 1865610806. E-mail add ess: [email p o ec ed]
&The Au ho 2015. Published by Ox o d Uni e si y P ess on behal o he Eu opean Socie y o Ca diology.
This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed
euse, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed.
Eu opace (2016) 18, 1287–1298
doi:10.1093/eu opace/eu 320
Downloaded om h ps://academic.oup.com/eu opace/a icle/18/9/1287/2802553 by Uni e sidad de Za agoza use on 26 Feb ua y 2021
Mo i a ion
Bo h biomedical esea ch and clinical p ac ice ely on complex da-
ase s o he physiological and gene ic cha ac e iza ion o human
hea s in heal h and disease. The in o ma ion shaping ou knowledge
o human hea s is ob ained om a a ie y o echniques and mod-
els, including eco dings ob ained in i o in asi ely and non-
in asi ely, in ex i o issue and isola ed human adul ca diomyocy es
eco dings, and mo e ecen ly in i o using human s em-cell-de i ed
ca diomyocy es. Inc easing e idence sugges s ha non-human ani-
mal models may ha e limi ed abili y o p edic human in i o e ec s
due o impo an species di e ences be ween humans, dogs, guinea
pigs, and abbi s.
1–3
Thus, me hods i mly oo ed in unde s anding
physiology and pa hophysiology in humans a e clea ly needed.
Ad ances in imaging echnologies such as he mul iple modali ies
o ca diac magne ic esonance (CMR) a e combined wi h body su -
ace o in a-ca diac elec ophysiological eco dings o e alua e in
speci ic pa ien s he in i o s uc u al and unc ional implica ions o
ca diac disease. Recen p og ess in esea ch using human ca dio-
myocy es de i ed om induced plu ipo en s em cells p omises ex-
ci ing new de elopmen s as i allows he in i o cha ac e iza ion o
he pheno ype o ca diomyocy es o speci ic pa ien s, and he e o e
has he po en ial o in oducing he lexibili y o in i o me hodolo-
gies in pe sonalized medicine. Fu he mo e, ex i o issue om biop-
sies o om dono human hea s p o ides measu emen s o issue,
cellula , and ionic p ope ies o human adul ca diomyocy es in non-
diseased and diseased condi ions.
Each o hese ypes o human-based assays and measu emen s
p o ides us a single snapsho om one pe spec i e o a complex
se o a iables h ough bo h ime and spa ial dimensions. In u n,
his complex se o a iables is able o de ine and explain he my iad
o dynamic mechanisms and p ope ies ha unde lie he ac i i y o
human hea s in heal h and disease. Gi en he complexi y and a -
ie y o app oaches and eco dings, he e is now g owing ecogni ion
o he need o embed compu a ional me hods in ca dio ascula
medicine and science o analysis, in eg a ion, and p edic ion. Com-
pu a ional app oaches in biomedicine ange o e a a ie y o echni-
ques o signal, da a, and image analysis, bu also impo an ly
mul iscale modelling and simula ion. Toge he , hey p o ide a syne -
gis ic app oach o o ganize and augmen he in o ma ion ob ained
om expe imen al and clinical eco dings. The bene i s gained in-
clude he quan i a i e analysis and o ganized eassembly o mul i-
scale and mul imodali y da ase s o p obe, challenge and expand
ou knowledge o he complex and dynamic in e ac ions in ca diac
elec ophysiology. Ad ances in compu a ional ca diac elec ophysi-
ology we e ecen ly illus a ed in wo dedica ed special issues o
Eu opace 2014.
4,5
Fu he mo e, new ini ia i es such as he Comp e-
hensi e in i o P oa hy hmia Assay (CiPA) launched by he Uni ed
S a es Food and D ug Adminis a ion (FDA) ecognize he po en ial
o human-based in silico and in i o app oaches as a new pa adigm o
d ug sa e y assessmen .
6
On 17 Sep embe 2014, a Wo kshop on Compu a ional Ca dio-
ascula Science was hos ed a he Uni e si y o Ox o d wi h he
aim o c ea ing an in e na ional, in e disciplina y, and in e -sec o ial
o um o discuss cu en ends in compu a ional echnologies o
augmen ca dio ascula physiology, pha macology, and medicine,
and o p opose solu ions o he eplacemen , e inemen and
educ ion o animal expe imen a ion.
7
In i ed pa icipan s included
expe s in ca diology, compu e science, physiology, pha macology,
philosophy, and biomedical enginee ing om academia and indus y,
om 11 Uni e si ies and 12 companies, om UK, se e al coun ies
in Eu ope, USA, and Japan. In his pape , we aim a desc ibing he
main ideas discussed du ing he mee ing, a he han p o iding a
ho ough e iew o he li e a u e. Table 1summa izes he o ms
o human-based in i o,ex i o,in i o,andin silico expe imen s
and echniques discussed, which a e u he illus a ed in Figu e 1.
The main ideas highligh ed h ough he wo kshop we e he
ollowing:
†A shi owa ds human-based me hodologies in pha macology
and medicine is occu ing, spu ed by ad ances in new in silico,
in i o,in i o, and ex i o echniques and he inc easing acknowl-
edgemen o he limi a ions o animal models.
†Compu a ional app oaches complemen , expand, b idge, and in-
eg a e in i o,in i o, and ex i o expe imen al and clinical da a
and me hods, and as such hey a e an in eg al pa o human-
based me hodologies in pha macology and medicine.
†The e ec i e implemen a ion o mul i- and in e disciplina y ap-
p oaches, eams, and aining combining and in eg a ing compu-
a ional me hods wi h expe imen al and clinical app oaches
ac oss academia, indus y, and heal hca e se ings is a p io i y.
†The human-based c oss-disciplina y app oach equi es expe s in
speci ic me hodologies and domains, who also ha e he capaci y
o communica e and collabo a e ac oss disciplines and o wo k
p oduc i ely in in e disciplina y and c oss-sec o en i onmen s.
†This new ansla ional domain o human-based ca diology and
pha macology equi es new pa ne ships suppo ed inancially
and ins i u ionally ac oss pha ma and bio ech indus y, con ac
esea ch o ganiza ions (CRO), academia and esea ch ins i u es,
echnology and se ice p o ide s, non-p o i and go e nmen al
o ganiza ions, and egula o y agencies. Ins i u ional, o ganiza ion-
al, and social ba ie s mus be iden i ied, unde s ood, and o e -
come in each speci ic se ing.
Desc ip ion o he wo kshop
The wo kshop consis ed o ou sessions aiming a explo ing di e -
en aspec s o human-based ca dio ascula science, and speci ically
he syne gies wi h in silico app oaches in he h ee main expe imen-
al se ings, i.e. in i o,in i o,andex i o. Table 1summa izes he
o ms o expe imen s and echniques discussed.
Human in i o ca dio ascula science
P o . Pe e Tagga examined he challenges in ol ed in ob aining
basic elec ophysiological da a om in i o human subjec s, du ing
ou ine clinical p ocedu es. The example discussed was he applica-
ion o pe cu aneous ansluminal co ona y angioplas y (PTCA) o
s udy in i o he e ec o myoca dial ischaemia (a majo cause o
mo ali y) on human elec ophysiology, by eco ding ei he mono-
phasic ac ion po en ials o unipola elec og ams on he en icula
endoca dium in he egion se ed by he a e y unde going PTCA.
In i o elec ophysiological eco ding echniques include he use o
he mul i-elec ode sock (264 elec odes) o e he human en i-
cles o s udy global pa e ns o ac i a ion such as en icula
B. Rod iguez e al.1288
Downloaded om h ps://academic.oup.com/eu opace/a icle/18/9/1287/2802553 by Uni e sidad de Za agoza use on 26 Feb ua y 2021
ib illa ion, ac ion po en ial du a ion (APD) changes, and pos -
epola iza ion e ac o iness du ing ea ly ischaemia and he demon-
s a ion o mechano-elec ic eedback in humans.
10–20
Findings
add essed he deba e as o whe he o o s o mul iple wa ele s
d i e en icula ib illa ion (VF) and showed ha bo h coexis in
human VF, and he e y apid ime cou se o he ea ly elec ophysio-
logical changes du ing ea ly ischaemia in humans. In i o elec o-
physiological eco dings a e c i ical o ansla ional esea ch
( om basic o clinical) bu hey a e limi ed due o conside a ions
o pa ien com o and sa e y. The e o e, hey need o be
complemen ed by and combined wi h al e na i e ways o p obing
he human hea , including non-in asi e in i o imaging as well as
ex i o in es iga ions, as desc ibed below.
D Rina A iga p o ided an o e iew o mul imodali y magne ic
esonance imaging and i s applica ion in pa ien s wi h hype ophic
ca diomyopa hy (HCM). HCM is he mos common gene ic hea
disease (a ec s 1 in 500)
21
and he commones cause o sudden ca -
diac dea h in he young.
22
T ans ho acic echoca diog aphy (TTE) is
ou inely used o assess HCM, bu is limi ed in pa ien s wi h poo
acous ic windows o poo isualiza ion o some LV egions. CMR
...............................................................................................................................................................................
Table 1 Summa y o human-based in i o,ex i o, and in i o echniques and in silico app oaches
Human-based Da a acquisi ion echnique Limi a ions o he da a P og ess o in silico app oaches
In i o Elec ophysiological eco dings du ing
clinical p ocedu es
[Tagga , Zhou]
– Limi ed da a se s due o
e hical and p ac ical
obs acles;
– Da ase s de i ed om
diseased hea s;
– In e -pa ien a iabili y;
– Limi ed expe imen a ion
– Signal analysis and in eg a ion;
– Mul iscale elec ophysiological models and
simula ions;
– Popula ion o models o mimic ac ion po en ial
a iabili y
Mul imodali y imaging including
magne ic esonance
[A iga, G au]
– Limi ed da a se s;
– No expe imen a ion
– Image analysis ( issue cha ac e iza ion);
– Ven icula shape analysis;
– S uc u al models and simula ions
– Compu e models o link be ween s uc u e and
di usion;
– And link be ween mic o s uc u e and unc ion
Body su ace po en ials,
elec oca diog am
[Minchole, Lu]
– Cap u es global pa e ns
o hea beha iou ;
– Va iabili y
– Au oma ed quan i ica ion o ECG ea u es o clinical
diagnosis and iden i ica ion o new bioama ke s
(mo phological QRS o T-wa e based, iCEB);
– Elec oca diog aphic imaging;
– Mul iscale human hea simula ions om ion channel
and mic os uc u e o he elec oca diog am
mHeal h eco dings h ough mobile
de ices
[Os e ]
– Ve y la ge quan i ies o da a
no amenable o manual
analysis;
– Noisy and pa chy da a;
– Social, e hical and legal
challenges
– Au oma ed and semi-au oma ed echniques o
analysis, such as machine lea ning, implemen able on
mobile phones
Isola ed human p ima y cells and
non-clinical, eal-wo ld da a om
biopsies and medical his o ies
[Ghe i]
– Limi ed da a se s;
– Social, e hical and legal
challenges
– Compu a ional models o in eg a e expe imen al
da a and o in es iga e mul iscale mechanisms o
disease and pha macological in e en ions
Ex i o Mic oelec ode, op ical mapping, pa ch
clamp, p o ein, and mRNA
exp ession
[Va o, B i on, Du a]
– Limi ed a ailabili y, and
mos ly om diseased
hea s;
– Va iabili y;
– Di icul y o echnique
implemen a ion (cell
isola ion; cu en
sepa a ion);
– Change o p ope ies due
o cell isola ion
– Da a analysis and in eg a ion;
– Mul iscale models o g ea e con ex ualiza ion;
– In es iga ion o a iabili y h ough app oaches such
as popula ion o models
In i o Human ca diomyocy es de i ed om
induced plu ipo en s em cells
(hiPSC-CMs)
[Daniels, Se e i, Koplja , Ha me ]
– Inconsis en imma u i y;
– Va iabili y and associa ed
di icul y o compa ison
– Models o in es iga e a iabili y, assis in e p e a ion,
and acili a e ansla ion o in i o cells
– Models o in es iga e gene mu a ions;
– Models o d ug sa e y assessmen
Cell cul u es and high-speed op ical
imaging
[Bu on]
– Limi ed c oss-ins i u ion
and c oss-sec o access o
expe imen s
– Mul iscale modelling o explain dynamics in
he e ogeneous p epa a ions
Human-based app oaches o pha macology and ca diology 1289
Downloaded om h ps://academic.oup.com/eu opace/a icle/18/9/1287/2802553 by Uni e sidad de Za agoza use on 26 Feb ua y 2021
is now he gold s anda d in assessing mass, hype ophy, olume,
and unc ion in HCM due o high spa ial and empo al esolu ion.
23
Unlike TTE, CMR also p o ides issue cha ac e iza ion using la e
gadolinium enhancemen (de ec s ocal ib osis which has been as-
socia ed wi h en icula a hy hmias and SCD)
24
and T1 mapping
(de ec s di use and ocal ib osis). CMR also o e s insigh s in o se -
e al hallma k ea u es o HCM ha a e po en ial con ibu o s o dis-
ease p og ession using no el echniques such as s ess pe usion
imaging (impai ed pe usion),
25
blood oxygen le el dependen
(BOLD) imaging a s ess (deoxygena ion a s ess),
26
phospho us
MR spec oscopy (abno mal myoca dial ene ge ics a es wi h u -
he de ici in exe cise),
27
and mos ecen ly, di usion enso im-
aging ( o assess ib e disa ay).
28
CMR is no only a use ul imaging
adjunc in cases wi h limi ed TTE iews, bu also p o ides accu a e
disease cha ac e iza ion o sub le di e ences. Resea ch using CMR
is imp o ing ou unde s anding o his complex he e ogeneous dis-
ease and is helping o guide isk s a i ica ion and ea men
s a egies.
29
P o . Vicen e G au spoke abou me hods o in es iga e myoca -
dial mic os uc u e, including quan i ica ion using imaging as well as
de e mina ion o unc ional epe cussions o mic os uc u al
changes, in es iga ed using compu a ional modelling and simula ion.
Recen de elopmen s in MRI echnology, in pa icula using di usion
sensi i e sequences, allow he quan i ica ion o mic os uc u e, ini-
ially in ixed hea s
8,30,31
o in hea s a di e en s ages o con ac-
ion.
32
Di usion MRI uses an indi ec measu emen o es ima e
ca diac s uc u e, and i s ela ionship o ca diac mic os uc u e is
no ully unde s ood. Two me hods o imp o e his unde s anding
we e discussed. His ology o e s di ec insigh s in o mic os uc u e,
bu h ee-dimensional econs uc ion om his ological slices is
Body
Body su ace po en ials
ECG eco dings
mHeal h da a
A hy hmic- isk bioma ke s
High- esolu ion
a hy hmia mechanisms
A ial and en icula
elec ophysiology models
in heal h and disease
1
0.5
0
−0.50 200 400
Time (ms)
600
Po en ial
2.00
1.38
0.75
0.12
−0.50
O gan
Clinical EP S udies
Mul i-modali y clinical MRI
Tissue
High- esolu ion MRI
His ology
Op ical mapping
Cell cul u es
Single cell
Mic oelec ode eco dings
P o ein/mRNA exp ession
Op ical mapping
Ionic cu en
Vol age/pa ch clamp
Isola ed cells/hiPSC-CMs
Figu e 1 Sou ces o expe imen al da a in eg a ed in compu a ional models o human ca diac elec ophysiology, and applica ions in physiology.
Ionic cu en models a e cons uc ed mos ly based on ol age/pa ch clamp da a om ex i o and in i o p epa a ions. The in eg a ion o ionic
cu en models in single cell models, accoun ing o a iabili y in p o ein exp ession and disease emodelling, allows o he simula ion o he ac ion
po en ial and elec oly e concen a ions in heal hy and disease. Addi ionally, ca diac simula ions a he whole o gan and body le els equi e he
cons uc ion o image-based ana omical models. When coupled o ma hema ical desc ip ions o elec ical exci a ion h ough ca diac issue, hey
allow o he high- esolu ion in es iga ion o a hy hmia mechanisms based on clinical elec ophysiology s udies, as o he in e p e a ion and
iden i ica ion o a hy hmic- isk bioma ke s a he su ace po en ial le el. T ansmu al isualiza ion o en icula myo ib e o ien a ion, adap ed
om e e ence 8wi h pe mission. ECG/whole body simula ion, adap ed om e e ence 9wi h pe mission.
B. Rod iguez e al.1290
Downloaded om h ps://academic.oup.com/eu opace/a icle/18/9/1287/2802553 by Uni e sidad de Za agoza use on 26 Feb ua y 2021
challenging.
33
Compu a ional models, simula ing wa e di usion and
MRI acquisi ion sequences, can be used o p o ide a di ec link be-
ween s uc u e and di usion.
34
The ela ionship be ween mic o-
s uc u al and unc ional changes is no ully unde s ood, and he e
compu a ional models can again p o ide a unique ool as shown
o example in e e ences.
35–37
In addi ion o imaging, clinical in i o eco dings o e alua e he hu-
man hea include he su ace elec oca diog ams (ECGs), a apid
and cos -e ec i e me hod o acqui e non-in asi e eco dings. A
la ge body o esea ch has been de o ed o he quan i ica ion o
ECG ea u es o pa ien s a i ica ion in e ms o a hy hmic isk
and disease.
38
D Ana Minchole
´desc ibed compu a ional ap-
p oaches applied o he ECG o he de ec ion o elec ophysio-
logical abno mali ies and pa ien s a i ica ion in HCM. Using a
da abase o Hol e eco dings ob ained in HCM pa ien s and olun-
ee s, she desc ibed he quan i ica ion o s anda d ECG-based bio-
ma ke s such as QT T peak o T end, as well as new ma hema ical
model-based mo phological ea u es o he QRS complex and
T wa e mo phology.
39,40
D Minchole also desc ibed how in o ma-
ion abou he elec ophysiological and s uc u al signa u e o dis-
ease om he ionic o he whole o gan le el can be in eg a ed in
mul iscale human hea models and used in he de elopmen and
unde s anding o ECG-based bioma ke s.
41
Mul iscale simula ions
as illus a ed in Figu e 2allow he iden i ica ion o key s uc u al
and unc ional ac o s ha de e mine each o he ECG bioma ke s
and p o ide a deepe and mo e p ecise unde s anding o he in o -
ma ion each o he bioma ke s con eys.
42
The new knowledge also
aids in he iden i ica ion o mo e selec i e and speci ic bioma ke s
o speci ic disease condi ions wi h complex unc ional and s uc-
u al signa u es such as HCM o myoca dial in a c ion. Body su ace
ECGs combined wi h imaging da a using ma hema ical algo i hms
o non-in asi ely econs uc he elec ical ac i i y on he epica dial
su ace o he hea in i o as demons a ed o example in
e e ences.
43–45
Human ex i o and in i o
ca dio ascula science
P o . And as Va o began by discussing he knowledge we ha e o
human elec ophysiology la gely h ough classic echniques such
as con en ional mic oelec ode and mo e mode n pa ch clamp
echniques, p o ein and RNA exp ession app oaches using human
ca diomyocy es om biopsies o dono hea s.
46–48
Ob aining
good ep esen a i e expe imen al human da a is es ic ed by a
numbe o p ac ical p oblems. Sou ces o issue pa icula ly en-
icula cells a e di icul o ob ain and isola ion o cells is compli-
ca ed by disease. The sou ce o undiseased dono hea s is
pa icula ly limi ed. The p ocess o cell isola ion may also lead o
changes om p ope ies s udied in mul icellula issue p epa a ions.
Fu he mo e, he sepa a ion o each o he ionic cu en s illus a ed
in Figu e 3is p oblema ic since he a ailable pha macological inhibi-
o s a e no o ally selec i e. Finally, P o . Va o highligh ed he im-
po ance o human-based s udies a he han highligh ed he use o
animal models. I is o en unde app ecia ed ha he e a e subs an-
ial species di e ences be ween human and e en la ge mammals,
such as he dog, conside ed o be a ep esen a i e model.
49,50
Fo
example, ma ked di e ences in en icula epola iza ion ese e
ha e been epo ed be ween human and dog,
49
wi h la ge apid
componen o he delayed ec i ying cu en (I
K
)inhumanbu
s onge slow componen (I
Ks
) and inwa d ec i ying cu en (I
K1
)
0.4–0.8
mV
I
0.226 0.3 0.4
gks (nS/pF)
0.5 0.586 –0.4
–1 1
0 0.8
Figu e 2 Compu e simula ion o he human hea elec ophysiology om ion channel o body su ace po en ials and he elec oca diog am.
Simula ions a e conduc ed using human biophysically de ailed models conside ing he e ogenei y in speci ic ionic p ope ies (le , colou scale co -
espond o he maximum conduc ance o he slow componen o he delayed ec i ying cu en ) o de e mine hei e ec on he spa io empo al
e olu ion o elec ical po en ials ac oss he whole o so (middle, ex acellula po en ials h oughou he o so) and on he ECG ( igh , main leads
displayed).
Human-based app oaches o pha macology and ca diology 1291
Downloaded om h ps://academic.oup.com/eu opace/a icle/18/9/1287/2802553 by Uni e sidad de Za agoza use on 26 Feb ua y 2021

in he dog. Consequen ly, APD p olonga ion caused by selec i e I
K
block is h ee- old la ge in human han in dog, which sugges s cau-
ion in ansla ion o animal indings o human. Species-speci ic di -
e ences in pha macological ac ion be ween animal models ha e
been also desc ibed,
50
wi h abbi exhibi ing la ge APD p olonga-
ion and p oneness o epola iza ion abno mali ies upon selec i e
I
K
block compa ed o o he s. Compu a ional modelling and simula-
ion can acili a e in e species compa ison by iden i ying and add es-
sing di e ences and maximizing he e-use o expe imen al da a
om human hea s.
D Oli e B i on desc ibed how he complex elec ophysio-
logical da ase s ob ained om ex i o human hea s ha e been in e-
g a ed in mul iscale compu e models, speci ically ocusing on
human da a o e he pas decade
51 –54
as examples. He desc ibed
he ecen cons uc ion o a popula ion o human en icula cell
compu a ional models ha cap u es he in e -subjec a iabili y
seen in ac ion po en ial eco dings om human en icula issue
p epa a ions om P o Va o’s labo a o y. The models in he popu-
la ion ha e a wide ange o di e en con igu a ions o ionic cu en
s eng hs o mimic a iabili y in a popula ion as in e e ence 55. The
eam in es iga ed whe he he e we e con igu a ions ha we e pa -
icula ly ulne able o de eloping epola iza ion abno mali ies such
as al e nans and ea ly a e depola iza ions (EADs), in esponse o
blockade o di e en combina ions o cu en s known o be impo -
an in epola iza ion— apid delayed ec i ying (I
K
), slow delayed
ec i ying (I
Ks
), inwa d ec i ying (I
K1
), and la e calcium (I
CaL
) cu en .
The compu a ional app oach he e o e in eg a es and ex ends ex-
pe imen al eco dings, gene a ing p edic ions and e ining hypo h-
eses abou p oa hy hmic mechanisms ha can hen be es ed
expe imen ally. Impo an ly, he in silico human models p o ide a
mul iscale amewo k o in es iga e wi h high spa io empo al eso-
lu ion key ionic mechanisms in d ug sa e y and e icacy in human,
wi h high deg ee o lexibili y in he possible in e en ions (such
as hea a e, concen a ions, and ad ene gic challenge) compa ed
o expe imen s.
P o . Ma Daniels e iewed he s a us o human ca diomyocy es
de i ed om induced plu ipo en s em cells (hiPSC-CMs). He saw
h ee uses o hese cells:
†sc eening o ca dio oxici y o no el he apeu ics,
†modelling and d ug disco e y o inhe i ed ca diac condi ions and
†in he medium o long e m, egene a i e cell he apy.
While he e is p og ess in scaling up p oduc ion o hese cells,
56
unc ional ma u i y is s ill a majo impedimen . S em cell de i a i es
in many ways ( o ce gene a ion, gene exp ession, sa come ic ma u -
i y, elec ophysiological p ope ies, e c.) esemble imma u e neo-
na al ca diac subs a es.
57
Howe e , he e is one key di e ence
ha is a he unique o hese cells and bes illus a ed by compa i-
son o he exis ing al e na i es. Adul ca diomyocy es a e consis -
en ly ma u e; neona al ca diomyocy es a e consis en ly imma u e.
By con as , s em cell de i a i es display an inconsis en imma u i y
(Figu e 4) such ha in any expe imen al es o da e, a iabili y wi hin
he sample is ypically one o de o magni ude o measu es such as
APD and cycle leng h as shown in e e ence 58. This will complica e
compa isons be ween samples, which a e ypically made on small
numbe s o cells.
D S e ano Se e i showed how a compu a ional app oach can be
suppo i e and complemen a y o he unc ional in i o s udy o
hiPSC-CMs. In exploi ing hiPSC-CMs as in i o models o he elec-
ophysiological e ec s o e ol ing o new d ugs, a de ailed unde -
s anding o he elec ophysiological p ope ies o hiPSC-CMs is
necessa y. In silico models o hiPSC-CMs ac ion po en ial ha e
been cons uc ed o con ol
60
and some gene ic mu a ions, such
as hose causing long QT (LQT) synd ome ype 1 (LQT1),
61
ype
2, and ype 3,
62
based on ecen elec ophysiological measu e-
men s
63 –66
and alida ed agains d ug adminis a ion. Compu e si-
mula ions showed ha in p inciple hiPSC-CMs a e quali a i ely
consis en wi h adul CMs in esponse o many cu en blocke s,
67
bu di e ences also eme ged. Mo eo e , hiPSC-CMs show a highly
a iable elec ophysiological beha iou , namely a a iable and depo-
la ized es ing po en ial and di e se a es o spon aneous ac ion
50
0
–50
V (mV)
INa (A/F)
ICaL (A/F)
IK1
IK
IKs
INaCa
INaK
I o (A/F)
I (A/F) I (A/F)
–100
–100
–200
–300
–1
–2
–0.2
1
0.5
0
1
0.5
0.2
0
0
0
0
0 0.5 1
0 0.5 1
0 0.5 1
0 0.5 1
0 0.5 1
0 0.5
Time (s)
1
Figu e 3 Simula ion o he human en icula ac ion po en ial
and he unde lying ionic cu en s. F om op o bo om, ime
cou se o he ac ion po en ial, sodium cu en (I
Na
), L- ype calcium
cu en , he apid and slow componen o he delayed ec i ying
cu en (I
K
,I
Ks
) and he inwa d ec i ying cu en (I
K1
), he ansi-
en ou wa d cu en (I
o
) and he sodium po assium pump (I
NaK
)
and he sodium calcium exchange (I
NaCa
).
B. Rod iguez e al.1292
Downloaded om h ps://academic.oup.com/eu opace/a icle/18/9/1287/2802553 by Uni e sidad de Za agoza use on 26 Feb ua y 2021
po en ials. Such high a iabili y can be pe cei ed as a limi ing ac o
o he applica ion o he compu a ional app oach o hiPSC-CMs. In-
deed, compu a ional models o cell elec ophysiology a e usually
de eloped on he basis o he a e age alues o quan i ies measu ed
in in i o expe imen s. The unde lying idea is ha he p edic ions
ob ained wi h he model o he ‘a e age cell’ can apply o all he cells
wi h a ole ance ha is o he same o de o he a iabili y o he
da a on which he model is based. The e o e, high a iabili y in
he da a esul s in p edic ions wi h lowe eliabili y. One way o
o e come his limi a ion is o include he a iabili y wi hin he model
i sel (which is no longe a model o he ‘a e age cell’) in o de o
ake i in o accoun bo h in he in es iga ion o physiological me-
chanisms and in model-based p edic ions. The a o emen ioned
popula ion o models app oach is a ele an and p omising example.
Fu he mo e, when expe imen al da a show high a iabili y, com-
pu a ional models can help o iden i y he causes o such he e o-
genei y and po en ially help o educe i . As a ele an example,
compu a ional analysis can be used o assess o wha ex en he a i-
abili y is due o (i) di e en expe imen al sys ems (e.g. cell lines) ha
could hen be desc ibed by di e en , speci ic, compu a ional models
(ii) di e ences in he exp ession o ionic channels om cell o cell,
which is e en ually much la ge han in adul ca diomyocy es o (iii)
lack o obus ness in he elec ical ac i i y o incomple ely ma u e
hiPSC-CMs. An example o he la e case a e he e ec s o small
changes in depola izing cu en s, which can lead o d ama ic
changes in he a e o spon aneous bea ing in hiPSC-CMs, whe eas
hey lead o only mino changes in es ing po en ial in adul ca dio-
myocy es. In his sense, he a iabili y can be educed i obse ed in
he model pa ame e ’s space since cells showing e y di e en APs
could be qui e close in e ms o hei ionic cu en maximal
conduc ances.
D I an Koplja con inued he heme looking a expe imen al
s udies on hiPS-CMs in sa e y assessmen s in he d ug de elopmen
se ing, wi h an emphasis on he in es iga ion o long- e m d ug
Va iable exp ession
MERGEAc ininT oponin I
Uno ganized Rudimen a y O ganized
Va iable sa come e ma u a ion
Figu e 4 S em-cell-de i ed ca diomyocy es ha e a iable pheno ypes: cu en me hods o s em cell di e en ia ion p oduce mixed popula ions
a wo dis inc le els—gene exp ession, and pos - ansc ip ion. This is demons a ed o sa come ic mo phology he e, wi h he panel on he le
showing wo cells in he ield o iew posi i e o he z-disc ma ke alpha-ac inin (whi e), and he hin ilamen p o ein oponin I (magen a). How-
e e , u he he e ogenei y exis s e en wi hin he cell popula ions exp essing bo h ma ke s, as only some cells demons a e o de ed sa come ic
uni s wi h clea cell pola i y (panel on he igh ). Me hods o elimina e (o compensa e o ) his will be needed o enable small di e ences be ween
samples o be iden i ied abo e he noise o he di e ence wi hin he sample. Scale ba 10 mm. The human ES line OXF2 was g own o con luency
on Ma igel and di e en ia ed as desc ibed in e e ence 59. Cells we e dissocia ed by incuba ion wi h ypsin/EDTA (0.05%, Li e ech) o 15 min a
oom empe a u e p io o seeding on o 0.1% gela in coa ed glass co e slips. Ten days a e seeding, cells we e ixed in 4% PFA (10 min, oom
empe a u e), pe meabilized (0.1% T i on X-100 in T is-bu e ed saline), and blocked wi h 2% BSA plus 0.001% sodium azide in TBS-T (1 h RT)
and incuba ed wi h P ima y an ibodies (mouse monoclonal an i alpha-ac inin, (sigma), and abbi polyclonal an i- oponin T, p io o washing and
incuba ion wi h Fab agmen an i-mouse 488, and an i- abbi 568 (molecula p obes). Images we e acqui ed on an up igh Leica SP5 con ocal wi h
a63×lens.
Human-based app oaches o pha macology and ca diology 1293
Downloaded om h ps://academic.oup.com/eu opace/a icle/18/9/1287/2802553 by Uni e sidad de Za agoza use on 26 Feb ua y 2021
e ec s. Cu en ly, a ious echnologies such as mul i-elec ode a -
ay (MEA), Ca
2+
ansien , impedance, and op ical ac ion po en ial
measu emen s a e applied on hiPSC-CMs o hei cha ac e iza ion.
Thei po en ial has been highligh ed by he CiPA ini ia i e launched
by he FDA. On he o he hand, d ug-induced delayed and ch onic
ca dio oxici y is one o he main isk o d ug wi hd awal om he
ma ke . The e o e, D Koplja desc ibed in es iga ions o he acu e
and delayed (5 days) e ec s o a ious oncological compounds on
hiPS-CMs using an impedance-based unc ional assay. Di e en
unc ional pa ame e s such as bea a e, cell index and incidence
o a hy hmia-like e en s we e e alua ed, and indica e ha he
hiPS-CMs can be used o de ec di e en le els o acu e and ch onic
ca dio oxici y and could be aluable in d ug sa e y.
In ano he indus ial pe spec i e, D Najah Abi-Ge ges discussed
cu en ly employed s a egies in he pha maceu ical indus y o
ca dio oxici y sc eening. He ga e an o e iew o he issue poin ing
ou he high a i ion a e o new molecules du ing he i e a i e d ug
disco e y p ocess: a signi ican numbe o hese a e caused by ca -
dio oxici y.
68
He highligh ed he ex en o which some sa e new
d ugs a e no de eloped because o hese po en ial conce ns. The
new CiPA ini ia i e a emp s o add ess his issue.
6
CiPA p oposes a
pa adigm based on sc eening new molecules agains speci ic ca diac
ion channels combined wi h in eg a i e compu e modelling o p e-
dic he p oa hy hmic po en ial o new d ugs. This is subsequen ly
combined wi h non- oden and ea ly human clinical s udies o as-
sess d ug e ec s on QTc measu emen s. D Abi-Ge ges highligh ed
he need o u he de eloping obus and p edic i e in silico models
ha ep esen na i e myocy e physiology and he hea o heal hy
olun ee s and pa ien s. Such models will p edic acu e and ch onic
d ug e ec s wi h high p edic i e alue on ECG abno mali ies o he
han QT/p oa hy hmia, hea a e, con ac ili y, blood p essu e,
and ca diac s uc u e. D Abi-Ge ges concluded by ad oca ing
ha scien is s, modelle s, egula o s, and pha maceu ical indus y
a e cu en ly well posi ioned o shape how u u e ca diac modelling
would posi i ely impac d ug de elopmen .
A hy hmia mechanisms and bioma ke s
A ou h session p esen ed he use o expe imen al and compu a-
ional me hods o in es iga ions in o a hy hmia mechanisms and
bioma ke s. D S ephen Ha me con inued he heme o iPSC ech-
nologies, in his case o in es iga ions on mechanisms o disease
pa hogenesis in LQT1 using iPSC echnology o model he he edi-
a y ca diac a hy hmia synd ome. Bo h he e ologous (HEK293/
CHO-K1 cell exp ession sys ems) and iPSC-based cellula models
a e used o in es iga e he unde lying disease mechanisms, and a
compa ison is conduc ed o e alua e whe he he disease mechan-
isms a e simila in bo h cell ypes. Resul s show di e ences on I
Ks
channel unc ion and a icking in he e ologous sys ems o i e
LQT1 pa ien mu a ions. The esul s a e being modelled compu a-
ionally o in es iga e he implica ions o I
Ks
channel mu a ions on
ac ion po en ial du a ion and mo phology, and he pene ance o
he mu a ion in human cell popula ions such as hose desc ibed in
he p e ious session.
D Sa a Du a p esen ed a s a e-o - he-a mul iscale human
whole hea compu a ional amewo k o in es iga e a hy hmic e -
ec s o d ugs and disease, and speci ically hERG block in acu e myo-
ca dial ischaemia. The amewo k p esen ed he e consis s o a
human ana omically based model wi h biophysically de ailed e-
p esen a ion o memb ane kine ics including ionic cu en and con-
cen a ion dynamics, as well as ib e o ien a ion and en icula
he e ogenei y. The human mul iscale model is pa ame e ized using
expe imen al da a, om ol age clamp a he ionic le el o MRI scans
a he whole hea le el. They p o ide a mul iscale pla o m o dis-
sec and analyse speci ic ischaemic and a hy hmic p ocesses wi h
high spa io empo al esolu ion, which is no possible o ob ain
h ough expe imen s alone, especially in human. The new insigh s
p o ided by he human model could help in he design o u he ex-
pe imen al and clinical s udies o imp o e pa ien isk s a i ica ion,
as well as decisions abou d ug dose du ing managemen o an i-
a hy hmic he apy. This s udy
69
can be ex ended o explo e he
mechanisms unde di e en condi ions, such as a ying sizes and lo-
ca ions o he ischemic egion, di e en d ug compounds and hei
mul ichannel e ec s, and in e -subjec a iabili y in ionic cu en s
and epola iza ion pa e ns. The p esen a ion he e o e highligh ed
he powe o human mul iscale simula ions using ana omically based
hea models o in es iga e sa e y and e icacy o pha macological
ac ion in le hal disease condi ions.
Using a di e en compu a ional app oach, D Julien Os e p e-
sen ed ad ances on he compu a ional de ec ion o a hy hmia epi-
sodes in ECGs. The de elopmen o mobile echnologies (mobile
phones, able s, e c.) o heal h se ices (mHeal h) is cu en ly ap-
idly g owing o wo main easons: (i) cos educ ion and (ii) access
o esou ce-sca ce communi ies.
70
The impo ance o he au o-
ma ic o semi-au oma ic de ec ion o a hy hmias on ECG eco d-
ings was highligh ed in he p esen a ion, gi en he simplici y o da a
acquisi ions and he e o e he mul iplica ion o such da a. Manual
expe analysis would be mo e a bu den o he clinicians han
help o he diagnosis o ca dio ascula diseases. Machine-lea ning
app oaches ha e been demons a ed o be a powe ul ool o
such an analysis o se e al applica ions, such as a ial ib illa ion
(AF) episodes o en icula ec opic bea de ec ion. Many Hol e
so wa es al eady equi e elec ophysiologis s o labo a o y echni-
cians o anno a e bea s clus e ed oge he au oma ically. Llamedo
and Ma inez
71
ecen ly sugges ed a echnique eques ing expe la-
belling o he clus e s, ou pu ed based on bo h mo phological- and
hy hm-based ea u es. O he echniques o hy hm classi ica ion
o AF episodes de ec ion, based uniquely on he hea hy hm,
ha e also been p esen ed, and a e implemen ed in implan able
de ices.
72
D Os e ’s p esen a ion ocused on wo majo a hy hmic ypes:
AF and p ema u e en icula con ac ion (PVC). A machine-
lea ning app oach has ecen ly been p oposed, h ough he imple-
men a ion o a Suppo Vec o Machine
73
on a mobile phone.
74
PVC is ano he ype o en icula a hy hmia, iden i ied as a p e-
dic o o mo ali y a e myoca dial ischaemia.
75
The applica ion
o an ECG mo phology model-based Bayesian il e ing
76,77
was
shown o be e ec i e o PVC de ec ion.
78
Ms Xin Zhou p esen ed compu a ional in es iga ions in o he io-
nic mechanisms unde lying epola iza ion al e nans in a popula ion
o human en icula models calib a ed using in i o eco dings such
as hose p esen ed by P o . Tagga . P e ious esea ch in o ca diac
al e nans has mainly been ca ied ou in animals a he han in hu-
man. Ms Zhou p esen ed he cons uc ion and calib a ion o a popu-
la ion o o e 2000 human en icula cell elec ophysiology models
B. Rod iguez e al.1294
Downloaded om h ps://academic.oup.com/eu opace/a icle/18/9/1287/2802553 by Uni e sidad de Za agoza use on 26 Feb ua y 2021
o mimic he ac ion po en ial a iabili y exhibi ed in in i o elec o-
physiological eco dings om 41 pa ien s.
79
The in silico human en-
icula cell popula ion is illus a ed in Figu e 5and i was shown o
ep oduce wo ypes o al e nans es i u ion cu es also obse ed
in human in i o eco dings. By analysing he popula ion o human
in silico models, she dissec ed he mechanisms unde lying ca diac
al e nans and how he complex in e ac ion be ween sa colemmal
cu en s and calcium dynamics con ibu e o he ini ia ion and
main enance o each al e nans ype. The e o e, in his p esen a ion,
mul iscale human in silico models we e used o cap u e key epola -
iza ion p ope ies o in i o eco dings, o in es iga e he ionic
mechanisms unde lying he occu ence o p oa hy hmic epola i-
za ion al e nans and o iden i y po en ial an i-a hy hmic a ge s.
D Hua Rong Lu desc ibed a new, non-in asi e and ansla ional
bioma ke — he index o ca diac elec ophysiological balance
(iCEB, he a io be ween QT and QRS)–in d ug and ischaemia-
induced ca diac a hy hmias and in gene ic LQT synd ome and B u-
gada synd ome. Cu en ly used bioma ke s may no be adequa e o
de ec all ypes o d ug-induced ca diac a hy hmias. Fu he mo e,
he e is also a need o a new bioma ke o de ec ca diac isks in
pa ien s wi h gene-de ec s in he hea such as LQT synd ome
and B ugada synd ome, in pa ien s wi h hea ailu e and o de ec
ca diac isks in spo smen and women. iCEB was success ully
de eloped and in oduced in 2013,
80
i may de ec po en ial isks
o d ug-induced ca diac a hy hmias beyond LQT and To sade
de Poin es. I may also a be be e han cu en ly used bioma ke s
de i ed om animal models, such as he isola ed abbi le -
en icula wedge model, because iCEB is also de ec ing addi ional
d ug-induced po en ial ca diac a hy hmias by slowing conduc ion
and QT-sho ening.
80
iCEB was also ound o be signi ican ly in-
c eased in pa ien s in gene ic LQT synd ome and signi ican ly de-
c eased in pa ien s wi h B ugada synd ome.
New pe spec i es/beyond he human
hea
D Jean-Pie e Valen in in oduced he backg ound o compound
es ing s a egies in esponse o ICH guidelines and how he CiPA
ini ia i e
6,81 –83
will de elop a new non-clinical pa adigm o ca diac
sa e y e alua ion o new d ugs by shi ing he ocus away om QT
p olonga ion o an assessmen o p oa hy hmia o mi iga e agains
he ho ough QT s udy. The p oposals would be o conside using
ion channel e ec s in andem wi h u ilizing he eme ging echnology
o in silico assessmen and s em-cell-de i ed ca diomyocy e e ec s,
bu no a he expense o clinical ECGs o an unde s anding o pha ma-
cokine ics/pha macodynamics.
6
The ini ia i e aims o deli e an imple-
men a ion o ecommenda ions ini ially by mid-2016 onwa ds;
howe e , his is dependen on communi y suppo in unde s anding
mo e abou which expe imen al inpu s, which models and which le els
o p edic i e capaci y a e equi ed.
6
Finally, Valen in eminded he audi-
ence o a need o keep QT in pe spec i e and o conside ca diac e -
ec sbeyondQTasi onlyaccoun s o 4%o ca diacsa e y- ela ed
d ug a i ion.
84,85
Among he ad e se e en s ha a e obse ed a e a -
hy hmia, achyca dia and changes in blood p essu e.
84,85
D And e Ghe i se ou he p oblem acing he de elopmen o
pha maceu ical d ugs in he ansla ion be ween in i o and animal
s udies o he clinical se ing and he c i ical need o ha e s onge
models a he non-clinical phase o imp o ing success and unde -
s anding o d ug ac ion. The app oach ha D Ghe i’s company Ana-
Bios will be aking is o use p ima y human issue o de i e cells ha
can be used mo e eliably in p edic ing la e clinical e ec s han in i o
o animal s udies can alone. This wo k is po en ially o be used in pa -
allel o he de elopmen o in silico app oaches ha a e be e in-
o med (pa ame e ized) by he da a being gene a ed om hese
isola ed human p ima y cells. Taken oge he wi h a much imp o ed
cha ac e iza ion o dono s, including medical his o y, allows us o con-
side he concep o inco po a ing eal-wo ld- ype da a in o models.
D Rebecca Bu on p esen ed he ecen wo k on de eloping a
cell cul u e model o neu ally media ed a hy hmogenesis and
non-in asi e op ical imaging me hods being pu sued a Ox o d.
86,87
Biological models wi h a ying deg ees o complexi y ha e been
de eloped o shed ligh on e-en an a hy hmias and ca diac mono-
laye s a e one o he simples models. The nex s ages o his esea ch
is o pu sue emo e moni o ing o in i o cell cul u es ha would in-
c ease expe imen al access, educe he need o sac i ice addi ional
animals, and spu he adop ion by o he labo a o ies wo king in allied
esea ch a eas. The e has been a p oli e a ion o emo e access pla -
o ms ha o e p omising unc ionali y. Remo e access imaging o e s
ad an ages o bo h ‘we -lab’ and ‘d y’ expe imen a ion (compu a-
ional and ma hema ical modelling), and he e is a need o pla o ms
which o e secu e and eliable implemen a ion.
Key challenges mo ing o wa d
In o de o succeed in a p og amme o esea ch and implemen a ion
o d ug disco e y and es ing ha akes ull ad an age o he s a e o
he a in ca dio ascula science, he wo kshop discussions iden i-
ied i e key challenges o be me
(1) Each o he human-based me hodologies and echniques p e-
sen ed a he wo kshop has ad an ages and disad an ages, as
80
Accep ed models
Rejec ed models
O iginal O’Ha a–Rudy model
40
0
–40
Memb ane ol age (mV)
–80
0 600
Time (ms)
Figu e 5 Popula ion o human en icula ac ion po en ial mod-
els calib a ed using in i o elec og am eco dings. Each simula ed
ac ion po en ial in he popula ion is gene a ed using he O’Ha a–
Rudy model wi h ionic conduc ances sampled in a wide ange o
possible alues. Calib a ion is hen conduc ed using he in i o elec-
og ams by selec ing he models ha yield ac ion po en ials wi h
p ope ies such as ac ion po en ial du a ion consis en wi h he
elec og ams ( ed aces, accep ed models), and ejec ing hose
ha a e ou side ange. In his igu e, he ac ion po en ial o he
o iginal O’Ha a–Rudy model is shown in black.
Human-based app oaches o pha macology and ca diology 1295
Downloaded om h ps://academic.oup.com/eu opace/a icle/18/9/1287/2802553 by Uni e sidad de Za agoza use on 26 Feb ua y 2021