Jou nal o In eg a i e Bioin o ma ics 2025; 22(1): 20250012
Danu a Schüle *, Ma hias Lange*, Thomas Al mann, Ma ia Cuacos, Daniel A end,
John Cha les D’Au ia, Anne Fiebig, Jochen Kumlehn, Ke s in Neumann, Michael Melze ,
Elena Rey-Mazón, Ha dy Rolle schek, Uwe Scholz, E elin Willne and Jochen C. Rei
Da a managemen in balance – a decade o
balancing p agma ism, sus ainabili y and
inno a ion a plan esea ch cen e IPK
Ga e sleben
h ps://doi.o g/10.1515/jib-2025-0012
Recei ed Feb ua y 14, 2025; accep ed Ma ch 25, 2025; published online May 30, 2025
Abs ac :The Leibniz Ins i u e o Plan Gene ics and C op Plan Resea ch (IPK) Ga e sleben is a leading in e -
na ional plan science ins i u e specializing in biodi e si y and c op plan pe o mance esea ch. O e he las
decade, all phases o he esea ch da a li ecycle we e implemen ed as a con inuous p ocess in conjunc ion wi h
in o ma ion echnology, s anda diza ion, and sus ainable esea ch da a managemen (RDM) p ocesses. Unde
he leade ship o a eam o da a s ewa ds, a esea ch da a in as uc u e, p ocess landscape, capaci y building,
and go e nance s uc u es we e success ully es ablished. As a esul , a gene ic esea ch da a in as uc u e was
c ea ed o se e he p inciples o good scien ific p ac ice, a chi ing esea ch da a in an accessible and sus ain-
able manne , e en be o e he FAIR c i e ia we e o mula ed. In his pape , we discuss success s o ies as well
as pi alls and summa ize he expe iences om 15 yea s o ope a ing a cen al RDM in as uc u e. We p esen
measu es o agile equi emen s enginee ing, echnical and o ganiza ional implemen a ion, go e nance, ain-
ing, and oll-ou . We show he benefi s o a pa icipa o y app oach ac oss all depa men s, pe sonnel oles, and
esea che p ofiles h ough pilo wo king g oups and da a managemen champions. As a esul , an ambidex-
ous app oach o da a managemen was implemen ed, e e ing o he abili y o e icien ly combine ope a ional
needs, suppo daily asks in compliance wi h he FAIR c i e ia, while emaining open o adop ing echnical
inno a ions in an agile manne .
Danu a Schüle and Ma hias Lange con ibu ed equally o his wo k.
*Co esponding au ho s: Ma hias Lange and Danu a Schüle , Leibniz Ins i u e o Plan Gene ics and C op Plan Resea ch
(IPK), D-06466 Ga e sleben, Ge many, E-mail: Ma [email protected] (M. Lange). schuele @ipk-ga e sleben.de (D. Schüle ).
h ps://o cid.o g/0000-0002-4316-078X (M. Lange). h ps://o cid.o g/0000-0003-4277-9879 (D. Schüle )
Thomas Al mann,Ma ia Cuacos,Daniel A end,John Cha les D’Au ia,Anne Fiebig,Jochen Kumlehn,Ke s in Neumann,Michael
Melze ,Elena Rey-Mazón,Ha dy Rolle schek,Uwe Scholz,E elin Willne and Jochen C. Rei , Leibniz Ins i u e o Plan Gene ics and
C op Plan Resea ch (IPK), D-06466 Ga e sleben, Ge many. E-mail: al mann@ipk-ga e sleben.de (T. Al mann),
cuacos@ipk-ga e sleben.de (M. Cuacos), a endd@ipk-ga e sleben.de (D. A end), dau ia@ipk-ga e sleben.de (J.C. D’Au ia),
iebig@ipk-ga e sleben.de (A. Fiebig), kumlehn@ipk-ga e sleben.de (J. Kumlehn), neumannk@ipk-ga e sleben.de (K. Neumann),
melze @ipk-ga e sleben.de (M. Melze ), mazon@ipk-ga e sleben.de (E. Rey-Mazón), olle @ipk-ga e sleben.de (H. Rolle schek),
scholz@ipk-ga e sleben.de (U. Scholz), willne @ipk-ga e sleben.de (E. Willne ), ei @ipk-ga e sleben.de (J.C. Rei ).
h ps://o cid.o g/0000-0002-3759-360X (T. Al mann). h ps://o cid.o g/0000-0003-4910-7311 (M. Cuacos).
h ps://o cid.o g/0000-0002-2455-5938 (D. A end). h ps://o cid.o g/0000-0002-4865-3938 (J.C. D’Au ia).
h ps://o cid.o g/0000-0003-3159-3593 (A. Fiebig). h ps://o cid.o g/0000-0001-7080-7983 (J. Kumlehn).
h ps://o cid.o g/0000-0001-7451-7086 (K. Neumann). h ps://o cid.o g/0000-0002-5213-4030 (M. Melze ).
h ps://o cid.o g/0000-0003-4813-5927 (E. Rey-Mazón). h ps://o cid.o g/0000-0002-8619-1391 (H. Rolle schek).
h ps://o cid.o g/0000-0001-6113-3518 (U. Scholz). h ps://o cid.o g/0000-0002-4153-4418 (E. Willne ).
h ps://o cid.o g/0000-0002-6742-265X (J.C. Rei )
Open Access. ©2025 he au ho (s), published by De G uy e . This wo k is licensed unde he C ea i e Commons A ibu ion 4.0 In e na ional
License.
2—D. Schüle e al.: A decade o da a managemen p ac ice a IPK Ga e sleben
Figu e 1: IPK oadmap o es ablish a esea ch da a managemen in as uc u e.
Keywo ds: esea ch da a managemen ; equi emen enginee ing; plan science; da a s ewa dship; LIMS; agile
da a flows and p ocesses
1 In oduc ion
The Leibniz Ins i u e o Plan Gene ics and C op Plan Resea ch (IPK) is a leading in e na ional plan science ins i-
u e wi h a esea ch ocus on biodi e si y and c op pe o mance. E ec i e esea ch da a managemen (RDM)
wi h he aim o c ea ing join ly usable da a spaces a ound he IPK genebank o plan gene ic esou ces is an
impo an basis o cu en and u u e inno a ions in basic esea ch, applied plan b eeding o o he conse a-
ion o biodi e si y. O e he pas decade, he IPK has ini ia ed i s digi al ans o ma ion p ocess. In subsequen
yea s, all phases o he esea ch da a li e cycle [1] and he associa ed FAIR p inciples [2] ha e been pu in o p ac-
ice as a con inuous p ocess in andem wi h in o ma ion echnology, s anda disa ion and sus ainable esea ch
da a managemen (RDM) p ocesses. The c oss-ins i u e RDM oadmap, as illus a ed in Figu e 1, s a ed in 2008
wi h a p ojec eam o ou c oss-depa men al esea ch g oups and headed by he Bioin o ma ics Uni o he
IPK. Commissioned by he boa d o di ec o s hey we e in cha ge o o mula ing a concep and oadmap o he
s a egic de elopmen o ins i u ional RDM.
In his pape we emba k on a jou ney o es ablish an ins i u ional RDM. We p esen measu es o agile
equi emen s enginee ing, echnical and o ganisa ional implemen a ion, i s go e nance, aining and oll-ou .
We discuss success s o ies as well as pi alls and summa ise he expe iences om 15 yea s o ope a ing a cen al
RDM in as uc u e.
2 Concep s udy o a gene al pu pose in o ma ion managemen
sys em
A p ojec eam was assembled in 2009, budge ed o e one yea and manda ed o conduc a s udy o p o ide
an objec i e basis o he decision-making p ocess. This s udy comp ised, an assessmen o exis ing p ac ice
o da a handling wi hin he ins i u e and a equi emen assessmen was conduc ed along wo ocus poin s:
D. Schüle e al.: A decade o da a managemen p ac ice a IPK Ga e sleben —3
echnical and ope a ional equi emen s. The echnical equi emen s included expandable da a s uc u e o
mapping s anda d labo a o y p ocesses, in ui i e, configu able use in e ace, mul ilingual capabili y, suppo
o s uc u ed and non-s uc u ed da a, connec ion o mobile de ices, audi ing, con olled ocabula ies, sea ch,
da a impo and da a expo in e aces and da a p o ec ion. Non- unc ional equi emen s we e sys em in e-
g a ion, expandabili y, in eg abili y in he o ganisa ional s uc u e, oll-ou model, a ailabili y and compliance
wi h da a secu i y egula ions.
The p ojec eam, which was in cha ge o elabo a e he s udy, was unde he umb ella o he Bioin o ma -
ics and In o ma ion Technology esea ch g oup, and comp ised as head a senio scien is wi h backg ound as
in o ma ion echnology enginee and wo doc o al bioin o ma ics s uden s, who we e unded o one yea , ep-
esen a i es o ou scien ific wo king g oups, known as pilo g oups, and he IPK’s Bioin o ma ics Coo dina o .
The pilo g oups we e selec ed o ep esen he ou depa men s o he IPK, o ensu e a high le el o in ol emen
in scien ific da a managemen p ac ices, e.g. by means o exis ing so wa e sys ems, li ed da a managemen p o-
cesses. In addi ion, when pu ing oge he he s udy eam, special ca e was aken o ensu e ha he equi emen s
o he indi idual depa men s we e co e ed as comp ehensi ely as possible, while a he same ime complying
wi h he pe o mance, sus ainabili y and unc ionali y demands o he in o ma ion managemen sys em o be
in oduced. The s udy1was handed o e in 2010. I compiled ecommenda ions and assessmen s on nine ocus
hemes [1,3]. An exce p is gi en below.
2.1 Pe sonal and o ganisa ional measu es
A key ecommenda ion o he s udy was ha he need o pool and e ain knowledge in o de o secu e he long-
e m in es men in a LIMS sys em should be eflec ed in he c ea ion o a sus ainable ole s uc u e. This should
be done (I) by c ea ing dedica ed job p ofiles o LIMS employees and (II) by ec ui ing and managing wi hin a
se ice subg oup wi hin an es ablished wo king g oup.
Fu he mo e, he oll-ou was also o be combined wi h he design o a aining p og amme. In he ea ly
days o aining, he wide ange o use s and aining equi emen s became appa en , which had o be adap ed
o he di e en needs and le els o knowledge o he espec i e wo k g oups and employees. Dedica ed aining
ocal poin s had o be se o he ollowing g oups in pa icula : PhD s uden s, scien is s and echnical s a .
A u he dimension was he specialised domain backg ounds ep esen ed a he IPK in plan biology, na u al
sciences and in o ma ion echnology.
Fo he in oduc ion, cus omisa ion, configu a ion, sys em in eg a ion and ope a ion o an IPK-LIMS, i
was ecommended ha he ollowing oles and wo k p io i ies be co e ed ei he by s a o be ec ui ed o
by syne gies wi h al eady exis ing s a :
Consul ing and aining – con inuous equi emen analysis; collec ion o da a managemen p ocesses, 1s le el
suppo .
So wa e enginee ing – ex ensions, expo and impo in e aces, de elopmen ailo ed on ends, 2nd le el
suppo .
Adminis a ion – moni o ing, issue managemen , so wa e upda es, configu a ion, use managemen , se e
managemen .
Managemen – cen al con ac poin LIMS and da a managemen issues; upda e and de elop esea ch da a
managemen concep s, ou each scien ific o p ojec s, esou ce esponsibili ies.
2.2 Cos s and expense es ima ion
The s udy highligh ed he s a egic e ec ha he in oduc ion o LIMS as a cen al se ice is likely o impac
on he s uc u e o he esea ch da a in as uc u e. The ollowing amewo k poin s we e he e o e se o a
esou ce es ima e o he sys em oll-ou :
1As he s udy con ains some sensi i e in o ma ion, i has no been published in ull. An exce p can be ob ained on eques .
4—D. Schüle e al.: A decade o da a managemen p ac ice a IPK Ga e sleben
Table 1: The equi ed RDM oles, he equi ed numbe o pe sonnel posi ions, es ima ed quali a i e esou ce e o o a LIMS oll-ou
and ope a ion.
Comme cial Open sou ce In-house
Rollou Ope a ion Rollou Ope a ion Rollou Ope a ion
Pe sonnel Da a s ewa d 2 1 2 1 1 1
So wa e enginee 1 1 2 1 3 2
IT adminis a o 0.5 0.25 1 0.5 0.5 0.25
Senio scien is 1 1 1 1 1 1
In es men equi emen s High High Low Middle Low Low
Ope a ing a ailabili y Low High Low Middle None Middle
Ope a ing expenses So wa e enginee ing High Low High High High High
Suppo Middle Middle Low High Low High
1. inc emen al in oduc ion in pilo g oups (up o wo yea s)
2. in eg a ion wi h IPK in o ma ion sys ems and da abases (one yea )
3. alloca ion o long- e m esou ces in he IPK budge and hei bundling in he IPK o ganisa ional cha
(pe manen )
4. con inuous de elopmen and main enance (pe manen )
5. in eg a ion in o he ins i u e’s aining p og amme (subsequen o he in oduc ion in he pilo g oups)
In addi ion o unc ional c i e ia, aspec s ela ing o pe sonnel and o ganisa ional measu es, he du a ion o
an in oduc ion and he main enance cos s incu ed in he long e m we e included in he e iew. In his con-
ex , comme cial sys ems, open sou ce sys ems and p op ie a y in-house de elopmen s we e compa ed. The
es ima ed wo kload and expenses include in es men s in pe sonnel and he numbe o posi ions equi ed o
he oles lis ed unde 2.6. as well as he in es men equi ed in so wa e, main enance and ope a ion (Table 1
– cos s and expense es ima ion).
The s udy was e alua ed by he boa d o di ec o s and led o he decision o in oduce a RDM in as uc u e
om a comme cial p o ide . The chosen so wa e endo is a specialis in LIMS sys em enginee ing (h ps://
www.limsophy.com/en), whose p oduc po olio includes an in eg a ed “Resea ch and Labo a o y In o ma ion
Managemen Sys em” (RALIMS) ha mee s all he equi emen s o mula ed and has a high ma ke p esence in
bo h public esea ch ins i u ions and p i a e companies.
The key aspec s in a ou o a comme cial endo we e he equi emen o long- e m sus ainable ope a-
ion, in es men sa ings, and he o al cos o owne ship. Especially in ligh o Open Sou ce e sus Closed Sou ce
deba e [4], he e we e p ima ily s ong a gumen s o ensu e compensa ions o pe sonnel fluc ua ions in e ms
o knowledge d ain, long- e m suppo o so wa e and sys em upda es, con inuous upda ing o in e aces o
ensu e echnical compa ibili y wi h da a collec ion p ocesses. The la e includes he echnical de elopmen o
ins umen s, senso s, plan pheno yping and geno yping acili ies, and con inuously upda ed sys em documen-
a ion and aining ma e ials.
Fu he mo e, he suppo con ac comp ises a pe manen ly dedica ed p ojec manage and so wa e engi-
nee on he endo -side. This suppo s knowledge dissemina ion, educes knowledge loss du ing s a u no e
and s eng hens he ins i u ional LIMS ope a ion eam o scale ou in case o inc eased s a ing needs, e.g. aca-
ion, sys em and scien ific ins umen upg ades, da a flow suppo o esea ch p ojec s e c. This inc eased
agili y was, as shown in Table 1, complemen ed by p edic able financial planning and was e en mo e cos e ec-
i e han long- e m financing o in-house s a which high po en ial o fluc ua ion. This expe ience was made
du ing he es ablishmen o he IPK bioin o ma ics in as uc u e, he genebank in o ma ion sys em and IT se -
ices be ween 2002 and 2008 as a esul o a ede al and s a e unding p og amme. He e a cen al combined
Bioin o ma ics and IT in as uc u es we e se up a company le el. The co esponding main enance, suppo
and consul ing con ac s in place and a e one pilla o con inuous and s able se ice ope a ion.
D. Schüle e al.: A decade o da a managemen p ac ice a IPK Ga e sleben —5
Figu e 2: Co e en i ies and ela ions o he RALIMS da abase s uc u e.
2.3 Technology and sys emin eg a ion
A he echnical le el, ou cha ac e is ics o he RDM in as uc u e we e conside ed. Fi s , uni e sali y, o man-
age expe imen al da a and me ada a, p ojec s, ins umen s, and labo a o y no ebooks. Second, in e ope abili y
wi h exis ing in-house IT in as uc u e, e.g. ORACLE da abase sys em, Mic oso Windows desk op so wa e
and compa ible file s o e. Thi d, capabili ies o an agnos ic suppo o da a flows and suppo o open o ma
compa ible bulk da a impo s. And ou h, he model o long- e m sus ainable se ice.
The ocus was on he sys em in eg a ing o a RALIMS in o IPK’s IT ecosys em ha comp ises (a) an ORA-
CLE ela ional da abase, (b) a hie a chical s o age managemen (HSM) sys em o a chi ing LIMS- e e enced
p ima y da a files and (c) a Mic oso Windows Se e Clus e o hos ing he RALIMS on -end as a desk op
clien agnos ic emo e desk op applica ion. The unde lying da a s uc u e o RALIMS is gene ic and simila o he
In es iga ion-S udy-Assay (ISA) concep [5]. As illus a ed in Figu e 2, his consis s o da a en i ies and a ibu es
ha model a la ge pa o he da a gene a ed in a esea ch ins i u e and a e implemen ed e icien ly as ables in
an RDBMS. Mo e de ails o he da a s uc u e was published in [5].
In comple ion o he ISA co e, he en i y-a ibu e- alue (EAV) model is applied, which is a ene able me hod
o ep esen ing a bi a y in o ma ion on an objec . Acco ding o he cu en s o ed da a, he ISA co e co e s
abou 80 % o he use cases and can be implemen ed e icien ly in well s o age and access op imized RDBMS
backends. Specifically, he IPK ORACLE RDBMs backend ea u es a obus ela ional s o age engine in la ge-
scale en i onmen s. As indus y s anda d, i ea u es in-build pe o mance op imisa ion echnology such as
pa i ioning, bi map index, que y ec o isa ion, in-memo y s uc u es que y, caches e c. To combine his ela-
ional model based-based s uc u es wi h no SQL elemen s, a ibu e alue ex ensions, la ge bina y objec s, da a
s eams, g aph da a s uc u es, ex e nal files o JSON and XML documen da a ypes a e suppo ed as well. The
suppo o hyb id da a s uc u es is he co e pilla and i s well op imised implemen a ion in ORACLE da abase
s ack enables o hos da a o any use case and ensu e scalabili y and e icien ope a ion o e millions o da a
poin s [6].Figu e 3 shows he cu en , sys em-in eg a ed a chi ec u e o he RALIMS esea ch da a in as uc u e
a decade a e i s ini ial deploymen .
O e he pas decade, IPK so wa e enginee s ha e de eloped complemen a y componen s such as B API
[7], a REST ul emo e applica ion p og amming in e aces and exposed SQL based in e ace o que y abula
da a [5], da abase s o ed p ocedu es o connec o he Da aCi e API [8] o min DOIs as pe manen unique and
globally esol able da a se iden ifie , and op ions o expo ing FDO-complian da ase s, such as an ISA-TAB
6—D. Schüle e al.: A decade o da a managemen p ac ice a IPK Ga e sleben
Figu e 3: Sys em in eg a ion a chi ec u e o he IPK RALIMS in 2024: The use on -end componen (U), he RALIMS da a managemen
so wa e (D), he s o age in as uc u es (S) and he da a expo and da a access in e aces (E) a e di ided in o a da a low ollowing he
FAIR p inciples (highligh ed in g een) and componen s adap ed o he needs o p op ie a y da a lows, such as sensi i e da a
(highligh ed in o ange). The lowe indices indica e he ins ance o he espec i e sys em componen ha ea u es speci ic unc ionali ies,
which a e mo e closely indica ed by he da a low a ow.
and hei publica ion, o example in EMBL BioSamples [9] and ENA [9] o e!DAL-PGP [10]. In addi ion, s uc-
u es o e e encing he con olled ocabula y in c oss domain on ologies [11], such as he NCBI axonomy and
plan on ology, and o mapping o plan specific me ada a s anda ds, such as MIAPPE [12], we e implemen ed.
Finally, a sys em in eg a ion wi h IPK genebank in o ma ion sys em [13] was implemen ed o ensu e ha monised
ma e ial and sample managemen .
3 Do e ailing wi h da a managemen o se ice and esea ch
p ocesses
The a o emen ioned sys em a chi ec u e se es wo majo classes o da a managemen p ocesses o he IPK.
The fi s ca ego y a e sole se ice p ocesses o cen ally managed ins umen s ha a e u ilized in esea ch
p ojec s. They ollow an ins i u ional ag eed p ocess o p ima y da a cap u e and a e ope a ed in an o de -
p ocessing manne by IPK financed pe manen s a . Examples a e da a acquisi ion p ocesses like he high-
h oughpu sequencing and pheno yping p ocesses [14] o unpublished in e nal se ice p ocesses like oo phe-
no yping in he hizo on sys em o IPK’s whe he simula ion acili y ‘PhenosSphe e’ and chemical managemen
as shown in Figu e 4.
Bo h comp ise (a) defined pe sonnel and o ganiza ional esponsibili ies including defined ansi ion poin s
be ween he labo a o ies, he scien is and he LIMS p ojec eam as well as (b) defined s anda d-complian and
machine-p ocessable da a o ma s, (c) manda o y me ada a s anda ds, and (d) p e ious defined da a publica ion
p ocess o sequence da a and o pheno yping da a.
An exempla pheno yping p ocess, implemen ed as a se ice p ocess in LIMS is he sco ing o plan ai s
in g een houses o on fields. He e da a cap u e using he sma phone app PhenoApp [14] is he s a o a LIMS
da a flow. The clea ly designed and easy- o-use app could be in eg a ed well in o he da a cap u e p ocess. LIMS
enables use s o c ea e inpu files and me hods o he e alua ion. Me hods ha ha e al eady been desc ibed
can be selec ed again and/o eused in a modified o m. Di e en geno ypes so called accessions om he oil and
D. Schüle e al.: A decade o da a managemen p ac ice a IPK Ga e sleben —7
Figu e 4: UML ac i i y diag am o IPK hizo on pheno yping (A) and chemicals managemen (B).
odde plan asso men s a e assessed in a ious ials. The esea ch da a is eco ded exclusi ely using he app.
This includes con inuous eco ding wi h sco ing alues o he linking o images wi h sco ing alues. Ano he
ad an age is he abili y o ake pho os di ec ly wi h he app o documen a ion pu poses.
Ano he a ea o applica ion o he IPK-LIMS conce ns he documen a ion o all wo k wi h gene ically mod-
ified o ganisms (GMOs). Documen a ion o GMOs is essen ial o achie e scien ific goals and o p omo e sa e y,
anspa ency and us in he esponsible use o bio echnology. In gene al, wo k wi h GMOs is subjec o s ic
con ol, egula ed by co esponding laws and con olled by s a e adminis a i e o ices. To ensu e he sa e y and
documen a ion o GMO wo k a IPK, he LIMS has a GMO module which can documen all GMO- ele an da a,
om gene a ion, s o age ( oom lis s), wo k ca ied ou (cul i a ion, ha es ) o he des uc ion o he co espond-
ing GMOs. Da a access is pe sonalised and p o ec ed, and en ies and changes a e aceable. Each p ojec leade
has ull access and da a en y igh s o his o he own (labo a o y) a ea, bu no o o he wo king g oups. The e
a e de ailed lis s wi h all ele an in o ma ion such as he ype o GMO and i s sa e y le el (S1 o S2 acco ding o
he Gene ic Enginee ing Sa e y O dinance), selec ion ma ke s, dono and ecipien o ganism (species), s o age
loca ion, pu pose o use in specific scien ific p ojec s and p ojec leade . In o ma ion abou specific GMOs can
be exchanged be ween wo king g oups. This s ep is a p e equisi e o ano he wo king g oup o gain access o
he co esponding GMO. The LIMS also allows he au oma ed c ea ion o documen a ion (‘annual epo s’) in
acco dance wi h he Ge man Gene ic Enginee ing Reco ding Ac . This ype o documen a ion a IPK has been
ully e alua ed and app o ed by he esponsible S a e Adminis a ion O ice in Halle/Saale. The GMO module in
LIMS also allows he o ganised s o age o documen s, le e s, oom plans, co espondence, e c. ha cha ac e ise
he espec i e p ojec a ea. A eposi o y o his kind would no be easible wi hou he secu i y ea u es p o ided
8—D. Schüle e al.: A decade o da a managemen p ac ice a IPK Ga e sleben
by a LIMS. I he e o e se es as a benchma k o o he ins i u ions ha wo k wi h GMOs. In summa y, he IPK
LIMS (1) mee s legal and egula o y equi emen s, (2) ensu es aceabili y and con ol, and (3) gua an ees he
IPK’s liabili y and esponsibili y owa ds he en i onmen and socie y.
In con as , da a flows in esea ch p ojec s need o be mo e agile and a e less igidly s uc u ed, eflec -
ing he na u e o inno a ion-d i en science. He e, he men ioned co e se ice p ocesses a e do e ailed wi h
he imme si e analy ics d i en knowledge gene a ion in esea ch p ojec s [18]. An example is BRIDGE [15] a
esea ch p ojec o he geno ypic and pheno ypic cha ac e isa ion o ba ley samples om Ge man Fede al Ex
si u Genebank o plan gene ic esou ces [13], a esea ch p ojec o he geno ypic and pheno ypic cha ac e isa-
ion o mo e han 22 housand ba ley accessions o he IPK genebank.
He e, he p e-defined RALIMS se ice p ocesses sequencing, seed managemen and sco ing p ocess we e
applied and in e wea ed o manage mo e han 48,000 samples om sequencing and cul i a ion wi h abou
776,000 da a poin s.
Such in e wea ing o sole se ices p ocesses and p ojec specific ones is a join ac i i y o p ojec and co e
se ice s a wi h a high demand o a e y close in e ac ion. The da a a e exposed ia SQL iews o he RALIMS
da a backend h ough a web po al [16]. These and o he p ojec s, wi h a o al o mo e han six million samples
and e aby es o da a, a e incuba o s o building he capaci y o p o ide FAIR RDM p ocesses o ne wo ks such
as he Eu opean li e-sciences in as uc u e o biological in o ma ion (ELIXIR) [17] o a na ional le el in he
Ge man Bioin o ma ics Ne wo k (de.NBI) [18] o he Na ional Resea ch Da a In as uc u e (NFDI) (h ps://www
.n di.de) in he conso ia, FAIRAg o [19] and NFDI4Biodi e si y [20].
The hi d ca ego y a e hyb id se ice p ocesses. Those sha e common s eps and da a s uc u es, bu a e
mo e agile and d i en by indi idual and p ojec se -ups. Examples a e he in eg a ion o Elec onic Lab No e-
book (ELN) documen a ion o a chi al o imaging, like mic oscopy. He e we ha e sha ed p ocess elemen s, like
documen a ion o expe imen al se -ups, measu e me hods, documen a ion o ma e ial and sample p epa a ion.
The documen a ion and sha ing o expe imen al esul s and used p ocessing and da a analysis pipelines need
o be suppo ed in a flexible less s ic way.
P ominen example a IPK a e mic oscopy and he complex me abolomics lab wo k flows. Fo example,
di e en mic oscopes p oduce a ying ypes and amoun s o images, wi h Ligh shee Fluo escence Mic oscopy
being a no able case. This echnique is ideal o long- e m li e-cell imaging and/o imaging o la ge samples, o en
gene a ing ela i ely ew bu ex emely la ge files, some exceeding one e aby e. Such s uc u ed da a cap u e
p ocesses ac oss se e al dozens o ins umen s [21] equi es a well-designed esea ch da a flow in o backend
s o age and he documen a ion o he measu ed objec and images aken. In o de o ensu e FAIR s o age and
handling, he ollowing s eps a e implemen ed. Image nomencla u e ollows a naming con en ion consis ing
o an image numbe , ollowed by da e and ime au oma ically s amped du ing acquisi ion, ep esen ing he
fi s unambiguous iden ifie . Gi en he la ge da a size, ini ially images a e s o ed locally du ing expe imen al
p ocedu es. Once decided ha he images a e o good quali y, hey a e ans e ed o a file om whe e images
will be ans e ed in o HSM, espec ing a use -defined olde hie a chy. A he same ime, me ada a associa ed
wi h each image is eco ded by manually adding en ies in o a dedica ed module wi hin LIMS c ea ed specifically
o his mic oscope. Which me ada a is eco ded was defined a e ou weeks o mic oscope use, and include
in o ma ion abou he use , e.g. name, cos cen e, sample, e.g. species, o gan, ansgenic unique GMO numbe in
he LIMS GMO documen a ion module, image-specific me ada a, e.g. ype o expe imen , fluo escence colou s
de ec ed and associa ed p o eins o s ains, and he file name and he file pa h in HSM. Upon en y c ea ion,
LIMS c ea es an unique iden ifie ha will be associa ed wi h he image. Only aw da a is s o ed, gi en he size
o he files, and ha p ocessed da a can be easily egene a ed. By e e ing o he acquisi ion da e in he LIMS
en y and/o in he image name, i is s aigh o wa d o e e o he co esponding en y in o he ELN. The e,
ex ended in o ma ion on he expe imen al se up, and on image p ocessing s eps a e documen ed. All in all, his
in eg a ed app oach le e ages LIMS as a cen al hub, ensu ing mic oscopic da a is managed in a FAIR manne
by combining modules o GMO, ELN, and imaging-specific da a.
Ano he example is he documen a ion o me abolomics da a in an elec onic labo a o y no ebook (ELN).
In con ex o such mo e semi-s uc u ed documen a ion, i is essen ial o ollow bes p ac ices o ensu e da a
in eg i y, ep oducibili y and compliance wi h FAIR p inciples. The ELN mus fi s ha e a use pe missions
D. Schüle e al.: A decade o da a managemen p ac ice a IPK Ga e sleben —9
Figu e 5: His o y o ac i i ies o a ha monised esea ch da a managemen using RALIMS a he Leibniz ins i u e o plan gene ics and
c op plan esea ch.
hie a chy o p o ec sensi i e da a. S anda dized me ada a fields and naming con en ions a e essen ial o main-
ain consis ency and acili a e da a e ie al [22]. Use s c ea e empla es based on expe imen al en ies ha
a e ailo ed o he ype o me hods used (i.e. GC-TOF MS da a s UPLC-TOF MS o UPLC-DAD/FLD). These em-
pla es include expe imen al design pa ame e s, hei p ocedu es, eagen s used, as well as sample p epa a ion,
equipmen u ilised, special obse a ions and in ended downs eam analysis p ocedu es and s a is ical es s. I is
impe a i e o sepa a e he aw da a om hose da a ha a e un h ough any analysis pipelines. All da a en ies
a e ime s amped and a ibu ed o hose esponsible o he unning o he ins umen s and analysis o he da a
in o de o main ain a clea audi ail. In ou expe ience, le e aging an ELN ha is accessible ins i u e-wide
enhances he collabo a ion and da a sha ing be ween and wi hin indi idual g oups. The me ada a augmen ed
files can hen also be used o downs eam epo ing in s anda d o ma s [23] and submission o he p ope
me abolomic eposi o y da abases, like GNPS, o Me aboLigh s [24].
4 Lessons lea ned om a decade o cen ally o ganised esea ch
da a managemen in as uc u e
The es ablishmen o a cen alised echnical in as uc u e o esea ch da a managemen and digi ally alid
documen a ion o scien ific expe imen s wi h he ins alla ion o he RALIMS echnology pla o m in 2011 was
he beginning o a p ocess o FAIR da a managemen a he IPK ha con inues o his day. Figu e 5 show he
ac ions and efinemen s o e a decade o align o he equi emen s o he mul idisciplina y esea ch landscape
a IPK in alignmen wi h he in e na ional RDM ecosys em.
These ac i i ies can be subdi ided in o h ee ca ego ies: (a) ac ions o embed he sys em in he labo a o y
and esea ch p ocesses, (b) he con inuous efinemen and supply o echnical ea u es and (c) aining p o-
g ammes. Subsequen ly, an exce p is gi en o he majo lessons lea n in mo e han a decade o LIMS-based
esea ch da a managemen and i s e ec o IPK’s sus ainable bu agile esea ch da a managemen in as uc-
u e a e discussed.
Cen alisa ion o RDM is linked o he need o a s ong c oss-depa men and g oup communica ion, e.g.
o es ablish bes p ac ices, s anda d ope a ing p ocedu es, and build confidence in he benefi s o cen alisa ion.
In his con ex i became appa en how impo an i is o do his in a pa icipa o y p ocess in a collabo a i e
de elopmen . The basis o he es ablishmen o a cen al RDM in as uc u e ac oss domains and o ganisa ional
s uc u es a e well-chosen pilo wo king g oups as seedlings o he s ep-by-s ep oll-ou in o de o achie e he
highes possible le el o accep ance among he majo i y o employees and o e come a ce ain scep icism and
ea o complex lea ning p ocesses. Specifically, i was beneficial o emphasize he added alue o daily wo k
and o p omo e us ing communica ion a eye le el h ough join wo kshops and us -building on a pe sonal
le el wi h a high deg ee o social compe ence in o de o discuss issues ac oss all hie a chies. One example o
how a echnical solu ion ha could be implemen ed a an ad hoc basis made daily ou ine wo k conside ably
easie was he launch o a cen alised in en o y o chemicals and haza dous subs ances in RALIMS. Thanks o
his in eg a ed ca alogue ac oss all labo a o ies, p e ious emails o all enqui ies we e no longe necessa y.
