METHODS
published: 20 Decembe 2016
doi: 10.3389/ molb.2016.00082
F on ie s in Molecula Biosciences | www. on ie sin.o g 1Decembe 2016 | Volume 3 | A icle 82
Edi ed by:
F ancesco Nicassio,
Is i u o I aliano di Tecnologia (IIT), I aly
Re iewed by:
As id Desi ee Haase,
Na ional Ins i u e o Diabe es and
Diges i e and Kidney Diseases (NIH),
USA
Eleono a Lusi o,
Is i u e o Molecula Oncology
Founda ion (IFOM), I aly
*Co espondence:
Sil ia Bo ini
[email p o ec ed]
Alessand o Muzzi
[email p o ec ed]
†P esen Add ess:
Sil ia Bo ini,
Cen e Médi e anéen de Médecine
Moléculai e, Ins i u Na ional de la
San é e de la Reche che Médicale
U1065, Nice, F ance;
Luca Fagnocchi,
Depa men o Epigene ics,
Fondazione Is i u o Nazionale di
Gene ica Molecola e “Romeo ed
En ica In e nizzi”, Milan, I aly
‡These au ho s ha e con ibu ed
equally o his wo k.
Special y sec ion:
This a icle was submi ed o
Ribonucleop o ein Ne wo ks,
a sec ion o he jou nal
F on ie s in Molecula Biosciences
Recei ed: 05 Sep embe 2016
Accep ed: 30 No embe 2016
Published: 20 Decembe 2016
Ci a ion:
Bo ini S, Del To dello E, Fagnocchi L,
Dona i C and Muzzi A (2016)
PIPE-chipSAD: A Pipeline o he
Analysis o High Densi y A ays o
Bac e ial T ansc ip omes.
F on . Mol. Biosci. 3:82.
doi: 10.3389/ molb.2016.00082
PIPE-chipSAD: A Pipeline o he
Analysis o High Densi y A ays o
Bac e ial T ansc ip omes
Sil ia Bo ini1*†, Elena Del To dello1‡, Luca Fagnocchi1 †‡, Claudio Dona i2and
Alessand o Muzzi1*
1GSK Vaccines S l, Siena, I aly, 2Compu a ional Biology Uni , Resea ch and Inno a ion Cen e, Fondazione Edmund Mach,
San Michele all’Adige, I aly
PIPE-chipSAD is a pipeline o bac e ial ansc ip ome s udies based on high-densi y
mic oa ay expe imen s. The main algo i hm chipSAD, in eg a es he analysis o he
hyb idiza ion signal wi h he genomic posi ion o p obes and iden i ies po ions o he
genome ansc ibing o mRNAs. The pipeline includes a p ocedu e, align-chipSAD,
o build a mul iple alignmen o ansc ip s o igina ing in he same locus in mul iple
expe imen s and p o ides a me hod o compa e mRNA exp ession ac oss di e en
condi ions. Finally, he pipeline includes anno-chipSAD a me hod o anno a e he
de ec ed ansc ip s in compa ison o he genome anno a ion. O e all, ou pipeline
allows ansc ip ional p o ile analysis o bo h coding and non-coding po ions o he
ch omosome in a single amewo k. Impo an ly, due o i s e sa ile cha ac e is ics, i
will be o wide applicabili y o analyse, no only mic oa ay signals, bu also da a om
o he high h oughpu echnologies such as RNA-sequencing. The cu en PIPE-chipSAD
implemen a ion is w i en in Py hon p og amming language and is eely a ailable a
h ps://gi hub.com/sil iamic oa ay/chipSAD.
Keywo ds: high densi y a ays, iling a ays, mic oa ays, ansc ip omes, code:py hon
INTRODUCTION
The apid de elopmen o new and high- h oughpu echnologies o conduc genome-wide s udies
has d ama ically inc eased he abili y o disco e new non-coding egula o y RNAs. Among he
se e al ansc ip ome-p o iling me hods (Zhang e al., 2014), high-densi y DNA iling mic oa ays
(Selinge e al., 2000) ha e been success ully applied o a a ie y o ansc ip ome s udies (Apa icio
e al., 2004; Be one e al., 2004; Webe e al., 2005; C aw o d e al., 2006; Liu, 2007; Heidenblad
e al., 2008).
The analysis and in e p e a ion o high densi y mic oa ay da a is based on a p ecise de ini ion
o disc e e ansc ip ional uni s, hus equi ing a speci ic algo i hm o iden i y hem. In pa icula ,
he main challenge is o segmen he hyb idiza ion signal along he genomic coo dina e o
accu a ely ob ain ansc ip s bounda ies, especially when also non-coding egions a e p obed.
Di e en s a is ical algo i hms ha e been de eloped o p ocess high-densi y a ay da a and o
ob ain such segmen a ion. The wides exploi ed me hod was in oduced by Kampa e al. (Kampa
e al., 2004) and was successi ely implemen ed in he Tiling A ay So wa e (TAS) (h p://www.
a yme ix.com/es o e/pa ne s_p og ams/p og ams/de elope /TilingA ayTools/index.a x). The
me hod was based on he gene a ion o a ansc ip ion map cons uc ed by collec ing neighbo
exp essed p obes, i.e., p obes wi h a smoo hed in ensi y abo e a gi en h eshold. TAS ex ended he
Bo ini e al. PIPE-chipSAD: A Pipeline o High-Densi y A ay Analysis
me hod o Kampa e al. by es ima ing he signi icance o
di e en ial exp ession using a Wilcoxon signed- ank es wi hin
local windows o a gi en wid h, cen e ed on each p obe. Mo e
ecen ly, he Model-based Analysis o Tiling-a ays (MAT) was
in oduced by Johnson e al. (Johnson e al., 2006). MAT
s anda dized he p obe signal alue h ough a model, elimina ing
he need o sample no maliza ion. MAT used an inno a i e
unc ion speci ically designed o sco e egions o ch oma in
immunop ecipi a ion (ChIP) en ichmen , which allowed obus
p- alue and alse disco e y a e calcula ions. Howe e , bo h
me hods we e no able o de ec sho ansc ip s. A solu ion
o he segmen a ion p oblem was also p oposed by Hube e al.
(Hube e al., 2006) hey used a change poin de ec ion algo i hm,
based on a dynamic p og amming app oach, ha de e mined he
global maximum o he log-likelihood o a piecewise cons an
model. This model p o ides good pe o mances, bu i doesn’
ake he p obes posi ion in o accoun , making da a analysis less
accu a e in pa ially co e ed genomes. Finally, Thomassen e al.
(Thomassen e al., 2009) desc ibed a new app oach o add ess he
p oblem o segmen a ion using a sliding and expanding window
unning along he genomic coo dina e. Howe e , he size o he
windows employed by Thomassen e al. could assume only h ee
alues making he app oach na owly applicable.
Despi e he a ailabili y o all hese algo i hms (Kampa e al.,
2004; Hube e al., 2006; Johnson e al., 2006; Thomassen
e al., 2009), a comp ehensi e a ailable pipeline o high-densi y
a ay da a analysis is s ill lacking, in pa icula o bac e ial
ansc ip omes o which an inc easing amoun o genome
anno a ions a e becoming mo e and mo e accessible (Land e al.,
2015; Loman and Pallen, 2015). Hence o h, a ou knowledge,
no exis ing ool is able o analyse mo e han one expe imen a
he same ime especially i also he non-coding po ions o he
genomes a e p obed.
To add ess hese needs, we ha e de eloped PIPE-chipSAD,
a pipeline o conduc high-densi y a ay da a analysis. The
main algo i hm is chipSAD (chip Signal A eas De ec o ),
which p o ides he segmen a ion o he hyb idiza ion signal
and de ines he bounda ies o he de ec ed ansc ip s.
Anno-chipSAD pe o ms an anno a ion o hese egions
guiding o he iden i ica ion o new a chi ec u al ea u es as
ope ons, small-RNAs and an isense messenge RNAs. Finally,
align-chipSAD iden i ies he ansc ip ional uni s analyzing
mul iple expe imen s om di e en chip layou s a he
same ime.
He ein we p o ide explica ions o he main s eps o PIPE-
chipSAD and de ails o he algo i hms s a egies. We also
show he applica ion o PIPE-chipSAD on wo p e iously
published da ase s (Mellin e al., 2010; Fagnocchi e al.,
2015) used o s udy he ansc ip ome a ia ion o Neisse ia
meningi idis 1h q mu an s ain o show he abili y o PIPE-
chipSAD o handle expe imen s wi h di e en expe imen al
designs.
Abb e ia ions: CPR, co ela ed p obe egion, ie clus e s o con iguous p obes
ha ing a co ela ed signal; SAS, signal a eas, ie pu a i e ansc ibed egions; ORF,
open eading ame; UTR, un ansla ed ansc ibed egion.
MATERIALS AND METHODS
PIPE-chipSAD was designed o in es iga e wo colo mic oa ay
da ase s, bu i is sui able o analyse da a om di e en sou ces
(one colo mic oa ay o sequencing echnologies such as RNA-
seq) wi h minimal da a elabo a ion.
The pipeline is composed by i e s eps, as indica ed in he
low cha shown in Figu e 1, ha co espond o h ee main
p og ams. This module s uc u e p o ides mo e lexibili y in
he da a analysis because he use can access o he di e en
p og ams independen ly. The pipeline is eely a ailable a :
h ps://gi hub.com/sil iamic oa ay/chipSAD.
Iden i ica ion o T ansc ip ional Uni s
chipSAD is he i s p og am o he pipeline. The main algo i hm
de ines pai s o con iguous sliding windows along he genomic
coo dina e x. The wid h o he windows is i e a i ely inc eased o
include con iguous p obes wi h a consis en signal. Finally, wo
co ela ed p obe egions (CPRs) o di e en sizes a e posi ioned
a he wo sides o each coo dina e on he genome (Figu e 1A in
se 1 ( ed boxes)).
Then, i ollows he iden i ica ion o posi ions whe e he
M alue signal is subjec o an ab up change o in ensi y. A
gene alized (x)pa ame e o compa e he igh and le CPR
was de ined as in (Tushe e al., 2001) in o de o es ablish he
signal “change poin s.” To c ea e a signal a ea (SAS) a s a egy
based on he in ensi y alue o con iguous CPRs was employed
(Figu es 1B,C). In case o iling p obes design, he e alua ion
o he - es cu e is enough o assess he s a s and he ends
o he SAS, as shown in Figu e 1B. On he con a y, in case o
non- iling a ays (i.e., non uni o m densi y o p obes) shown
in Figu e 1C, he analysis o he - es cu e is no su icien
o p o ide good bounda ies assessmen hus he p ocedu e is
en iched wi h a c i e ion based on he M alue compa ison
and a signal smoo hing calcula ion based on he pseudomedian
o Hodges-Lehmann es ima o (Royce e al., 2007). B ie ly, he
a e aged in ensi y o wo egions, each limi ed by wo consecu i e
change poin s, was compa ed: i hei pseudomedian in ensi y
was g ea e han he absolu e alue o he di e ence be ween he
pseudomedian in ensi ies o he wo egions, hey we e joined
and he compa ison would ex end o he nex egion, o he wise
wo di e en SAS we e c ea ed. The esul o his s ep is he lis o
SAS o be in e p e ed as pu a i e ansc ip s.
Compa a i e Analysis o Mul iple
Expe imen s
In case o mul iple expe imen s, align-chipSAD can be un. In
o de o a oid ha noisy a eas may in luence he esul s o
his p ocedu e, only he SAS wi h a minimum alue o he
M pseudomedian (M) migh be selec ed be o e unning align-
chipSAD. Then, g aphs based on he o e lap o he SAS a e buil
as indica ed in Figu e 1D by he o ange links. Fo each connec ed
g aph, he consensus bounda ies a e calcula ed conside ing a
weigh ed a e age o he bounda ies o he SAS belonging o he
g aph. In de ail, he weigh ed posi ion hxi(s a o end) o he
F on ie s in Molecula Biosciences | www. on ie sin.o g 2Decembe 2016 | Volume 3 | A icle 82
Bo ini e al. PIPE-chipSAD: A Pipeline o High-Densi y A ay Analysis
FIGURE 1 | De ails o he i e consecu i e s eps o he PIPE-chipSAD.chipSAD algo i hm s eps: (A) In he i s s ep, one pai o sliding and expanding
windows is used o c ea e wo egions g ouping consecu i e p obes o co ela ed signal in ensi y, named Co ela ed P obe Regions (CPRs). Pai s o consecu i e
CPRs a e compa ed in he second s ep, using a modi ied - es , o iden i y he posi ions in which he signal changes signi ican ly. In (B,C) he bounda ies o he signal
a eas (SAS) a e de e mined o iling and non- iling p obes design espec i ely, i.e., po ion o he ch omosome whose p obes show simila in ensi y. This s ep p o ides
a lis o SAS bounda ies. Align-chipSAD:(D) This s ep is op ional bu necessa y in case o a compa a i e analysis o mul iple expe imen s. The esul o his s ep is a
unique lis o SAS bounda ies o se e al expe imen s, ins ead o one lis o each expe imen . Anno-chipSAD: (E) The las s ep deals wi h he associa ion o he SAS
wi h he ch omosome-wide anno a ion, Basically, i compa es he iden i ied SAS wi h he gbk ile o he o ganism o in e es .
F on ie s in Molecula Biosciences | www. on ie sin.o g 3Decembe 2016 | Volume 3 | A icle 82
Bo ini e al. PIPE-chipSAD: A Pipeline o High-Densi y A ay Analysis
consensus SAS is:
hxi = Pipixi
Pipi
whe e pi=Mi
PiMi
wi h i=1, 2 ...,Nenume a es each SAS belonging o he
g aph.
Thus, o each connec ed g aph, a consensus SAS is
de e mined, and he conclusion o his s ep is a unique lis o
consensus SAS.
Pu a i e T ansc ip s Anno a ion
The inal s ep o he PIPE-chipSAD is he compa ison o he
de ec ed SAS wi h he genome anno a ion and he classi ica ion
o pu a i e ansc ip s when hey ep esen single open eading
ames (ORFs), en i e ope ons (polycis onic ansc ip s),
an isense RNAs, un ansla ed ansc ibed egions (UTRs) and
in e genic RNAs (Figu e 1E). This p ocedu e consis s in he
compa ison o he iden i ied SAS wi h he gbk ile o he o ganism
o in e es and is implemen ed by anno-chipSAD; compa ing he
iden i ied SAS wi h he gbk ile o he o ganism o in e es . The
i s s ep o he classi ica ion was he selec ion o SAS o e lapping
anno a ed ORFs: i he SAS o e lapped a minimum o 30% o
a single ORF leng h in he same s and hen i was classi ied
as ORF (magen a a ow in Figu e 1E), o he wise as an isense
(o ange a ow in Figu e 1E). I he SAS o e lapped wo o mo e
co-o ien ed ORFs in he sense s and (wi h a minimum o 30% o
each ORF leng h), hen i was classi ied as ope on (b own a ow
in Figu e 1E). I he SAS o e lapped wo di e en ly o ien ed
ORFs and he in e genic egion be ween he wo ORFs was less
han 30 bp, hen i was classi ied as o e lapping UTR (pu ple
a ow in Figu e 1E). I he SAS did no o e lap an ORF in bo h
s ands and i s leng h was less han 800 bases, i was classi ied as
in e genic RNA (da k g een a ow in Figu e 1E). Finally, i he
SAS o e lapped bo h an ORF and he lanking in e genic egion,
a 5′o 3′ends, hen i was classi ied as UTR (5′o 3′ espec i ely)
only i he gap be ween he ORF and he in e genic egion is less
han 30 bases (black a ow in Figu e 1E).
Da ase s
We analyzed wo p e iously published da ase s used o s udy
he ansc ip ome a ia ion o N. meningi idis 1h q mu an
s ain wi h espec o wild ype. The ansc ip ome da a o he
N. meningi idis MC58 h q null mu an (1h q) s ain g own in
GC medium desc ibed by Fagnocchi e al. (Fagnocchi e al.,
2015) and he da ase p esen ed by Mellin e al. (Mellin e al.,
2010), which examined he N. meningi idis MC58 h q null
mu an and he ela i e complemen ed s ain, g own in i on
deple ed (100 µg des e al ml−1) o eple e condi ions (100 µg
e ic ni a e ml−1), we e used in his s udy, o es he capabili y
o chipSAD o handle da a om di e en expe imen al designs.
The mic oa ays analyzed a e one o wo-colo hyb idiza ions.
In case o wo-colo expe imen s (Fagnocchi e al., 2015) a
compe i i e hyb idiza ion be ween he null mu an and he wild
ype s ains g ow h unde he same in i o g ow h condi ion was
pe o med. In case o one-colo expe imen s (Mellin e al., 2010)
an in silico compa ison o wild ype s ains g ow h was composed
du ing global no maliza ion. Th ee biological eplica es o each
expe imen al condi ion we e a ailable in each da ase . In o de
o make he da a compa able, we me ged he eplicas o he same
condi ion by a e aging he M alue o each p obe a e a global
no maliza ion o signals by using limma R package (Smy h and
Speed, 2003).
Pa ame e Es ima ion
Au oma ic pa ame e es ima ion was done by he analysis o he
dis ibu ion o p obe M alues. The de e mina ion o a sui able
ini ial window size wwas cons ained by he spacing o he
p obes along he ch omosome. Thus, he window size wwas
es ima ed calcula ing he a e age dis ance be ween consecu i e
p obes. The pa ame e m(w) was es ima ed by calcula ing he
minimum numbe o p obes ound in a window wi h size w.
RESULTS AND DISCUSSION
Compa a i e Analysis o he T ansc ip ome
o N. meningi idis 1h q Mu an in Di e en
G ow h Condi ions
We analyzed wo p e iously published da ase s (Mellin e al.,
2010; Fagnocchi e al., 2015) used o s udy he ansc ip ome
a ia ion o Neisse ia meningi idis 1h q mu an s ain o es
he capabili ies o chipSAD o manage di e en chip designs and
hyb idiza ions in a single analysis (Figu e 2).
In Table 1 SM is epo ed he numbe o di e en ially
exp essed ansc ip s o each genomic ca ego y ound by PIPE-
chipSAD a e he alignmen o he wo da ase s. The applica ion
o PIPE-chipSAD allowed he de ec ion o di e en ially exp essed
ansc ip s, especially o ansc ip s loca ed in in e genic
egions, in he i e samples wi hou a ec ing he esul s ob ained
unning chipSAD on a single sample. In o de o see how he
alignmen p ocedu e, implemen ed in align-chipSAD, a ec s he
SAS iden i ied by chipSAD, we independen ly an he chipSAD
algo i hm only on he 1h q mu an expe imen o Fagnocchi’s
da ase and on he all expe imen s oge he and we compa ed
he esul s. Among he 114 egions classi ied as single ORFs
de ec ed in he single un, 104 a e s ill p esen a e he alignmen :
84 o hose main ain he same classi ica ion, while 20 egions
a e ea anged in di e en kind o ansc ip s mainly as mRNA
wi h UTRs o ope ons. Ten ansc ip s a e no longe de ec ed
as di e en ially exp essed a e he applica ion o align-chipSAD.
A e y simila scena io is obse ed wi h ope ons and mRNAs
wi h UTRs. The 93 mRNAs wi h UTRs iden i ied in he analysis
o a single expe imen o e lap 115 ansc ip s (some mRNAs
wi h UTRs comp ise wo genes), 74 o hem a e s ill classi ied
as mRNAs wi h UTRs a e he alignmen p ocedu e, while
33 ansc ip s a e classi ied in a di e en class, mo e o en in
single ORFs han in ope ons. Finally, 100 ou o 123 ansc ip s
classi ied as ope ons belong o he same classi ica ion wi h he
single un and upon align-chipSAD applica ion. The in e genic
egions a e he mos a ec ed by he alignmen p ocedu e, because
hese egions usually a e sho and noisy. One hund ed wen y-
six in e genic egions ou o 177 su i e o he alignmen o
which 32 a e joined in longe ansc ip s such as mRNA wi h
UTR o ope on. Twen y-se en in e genic egions a e no longe
p esen mainly because hey had M alues close o he h eshold.
F on ie s in Molecula Biosciences | www. on ie sin.o g 4Decembe 2016 | Volume 3 | A icle 82
Bo ini e al. PIPE-chipSAD: A Pipeline o High-Densi y A ay Analysis
FIGURE 2 | An example o he alignmen p ocedu e o he de ec ed signal a eas (SAS). The i e expe imen s a e: 1h q mu an g ow h in GC medium (GC,
om Fagnocchi e al., 2015) and ou expe imen s om Mellin e al. (2010): he 1h q mu an and he complemen ed mu an (comp) g own unde i on deple ed (DEP)
o eple e (REP) condi ions. (A) The SAS iden i ied by chipSAD o he i e expe imen s be o e he un o he alignmen p ocedu e. (B) The aligned SAS.
F on ie s in Molecula Biosciences | www. on ie sin.o g 5Decembe 2016 | Volume 3 | A icle 82
Bo ini e al. PIPE-chipSAD: A Pipeline o High-Densi y A ay Analysis
The e o e, he main e ec o he alignmen p ocedu e is he
ea angemen o he egions in di e en classes espec o he
classi ica ion ob ained applying chipSAD on a single expe imen .
Anyway, a e he alignmen , only a small pe cen age o he
ansc ip s ha e a class change, meaning ha he bounda ies
o he egions de ec ed by chipSAD, e en conside ing only one
expe imen , a e highly conse ed. Mo eo e , only ew ansc ip s
a e no longe de ec ed a e he alignmen and mos o hem
had e y low M alue, meaning ha he align-chipSAD can
imp o e he de ec ion o signi ican egions emo ing noisy
egions.
In o de o compa e ou esul s wi h he panel o 132
genes di e en ially exp essed al eady published in Mellin e al.
(Mellin e al., 2010), we i s selec ed he ansc ip s speci ically
di e en ially exp essed in he 1h q mu an s wi h espec o
bo h he wild ype and also he complemen ed s ain. Thus,
we se up a h eshold on he M alue, selec ing only hose
ansc ip s wi h a minimal exp ession old change o 2 be ween
he 1h q o he complemen ed s wild ype s ain, in each g ow h
condi ion. This selec ion c i e ion led o he iden i ica ion o
90 ansc ip s, including 35 ORFs, 28 ope ons, 10 mRNA wi h
UTRs, 17 in e genic egions (Figu e 3), ha a e he mos eliably
and consis en ly H q-modula ed and i on-dependen ansc ip s.
No ewo hy, we ound ha 42 genes we e iden i ied by bo h
app oaches, howe e PIPE-chipSAD was able o de e mine
whe he hese ansc ip s we e o ganized in mo e complex
s uc u es such as ope ons o mRNA wi h UTRs. Fu he mo e,
we indi idua ed 31 p e iously uniden i ied H q-modula ed and
i on-dependen ansc ip s. Finally, ou app oach ound 17 new
in e genic egions o be speci ically de egula ed unde i on
s a a ion/abundance and in absence o H q, p o iding also
he speci ic bounda ies o he pu a i e ansc ip s. O e all ou
me hod allowed a be e unde s anding o he H q a ge ome.
Finally, he esul s ob ained by he applica ion o align-chipSAD
made he compa ison o ansc ip s iden i ied ac oss da ase s
much easie , allowing a compac iew such as a hea map
(Figu e 3).
We ha e p esen ed PIPE-chipSAD, a speci ically designed
pipeline o bac e ial ansc ip omic analysis o high-densi y
a ay da a. The wo k low is a icula ed in h ee main p og ams:
chipSAD,align-chipSAD and anno-chipSAD. We p e iously
used chipSAD, o analyse he ansc ip ome a ia ion o N.
meningi idis in a ime cou se adap a ion o human blood (Del
To dello e al., 2012) and in esponse o physiologically ele an
g ow h condi ions (Fagnocchi e al., 2015). These wo success ully
applica ions showed he widely applicabili y o his ool and
he eliabili y o he achie ed esul s. He e, he me hod was
imp o ed o achie e be e pe o mances in he segmen a ion o
he in ensi y signal o bo h uni o m and no uni o m p obe a ay
designs (i.e., iling and non- iling mic oa ay design). Mo eo e ,
we ha e showed ha align-chipSAD was able o manage se e al
expe imen s, wi h di e en expe imen al design, analyzing hem
simul aneously. Fu he mo e, we ha e demons a ed ha PIPE-
chipSAD allows acing and s udying he ansc ip ional p o ile
o bo h coding and non-coding po ions o he ch omosome in
a single amewo k. O e all, bea ing he e sa ile cha ac e is ics
o PIPE-chipSAD, we belie e ha i will be o wide applicabili y
FIGURE 3 | Hea map isualiza ion o he H q-modula ed ansc ip s.
Hea map isualiza ion o he op H q-modula ed ansc ip s acco ding o hei
di e en ial exp ession s. he wild ype s ain in each mic oa ay expe imen .
The ansc ip s comp ise o 35 ORFs, 28 ope ons, 10 UTRs, 17 in e genic
egions.
F on ie s in Molecula Biosciences | www. on ie sin.o g 6Decembe 2016 | Volume 3 | A icle 82
Bo ini e al. PIPE-chipSAD: A Pipeline o High-Densi y A ay Analysis
and i migh be easily applied o analyse da a om o he high-
h oughpu echnologies such as RNA-seq.
AUTHOR CONTRIBUTIONS
SB and AM designed and concei ed he s udy. SB de eloped
he so wa e and analyzed he da a. SB, ED, LF, CD, and
AM con ibu ed o de elopmen o he analysis ool and o
esul s in e p e a ion. SB and AM w o e he manusc ip . All
au ho s con ibu ed, p o ided commen s and app o ed he inal
manusc ip .
FUNDING
SB was ecipien o a No a is Vaccines ellowship om he Ph.D.
p og am o he Uni e si y o Siena. LF and ED we e ecipien
o a No a is Vaccines ellowship om he Ph.D. p og am
o he Uni e si y o Bologna. The au ho AM was employed
by he unding o ganiza ion, No a is Vaccines now a GSK
company.
ACKNOWLEDGMENTS
The au ho s would like o hank D. Se u o, I. Delany, and
M. De Chia a o help ul discussions and o help in edi ing he
manusc ip .
SUPPLEMENTARY MATERIAL
The Supplemen a y Ma e ial o his a icle can be ound
online a : h p://jou nal. on ie sin.o g/a icle/10.3389/ molb.
2016.00082/ ull#supplemen a y-ma e ial
REFERENCES
Apa icio, O., Geisbe g, J. V., and S uhl, K. (2004). Ch oma in
immunop ecipi a ion o de e mining he associa ion o p o eins wi h
speci ic genomic sequences in i o.Cu . P o oc. Cell Biol. Chap e 17, Uni 17.
7. doi: 10.1002/0471143030.cb1707s23
Be one, P., S olc, V., Royce, T. E., Rozowsky, J. S., U ban, A. E., Zhu, X., e al.
(2004). Global iden i ica ion o human ansc ibed sequences wi h genome
iling a ays. Science 306, 2242–2246. doi: 10.1126/science.1103388
C aw o d, G. E., Da is, S., Scache i, P. C., Renaud, G., Halawi, M. J., E dos,
M. R., e al. (2006). DNase-chip: a high- esolu ion me hod o iden i y
DNase I hype sensi i e si es using iled mic oa ays. Na Me hods 3, 503–509.
doi: 10.1038/nme h888
Del To dello, E., Bo ini, S., Muzzi, A., and Se u o, D. (2012). Analysis o he
egula ed ansc ip ome o Neisse ia meningi idis in human blood using a iling
a ay. J. Bac e iol. 194, 6217–6232. doi: 10.1128/JB.01055-12
Fagnocchi, L., Bo ini, S., Gol ie i, G., Fan appiè, L., Fe licca, F., An unes,
A., e al. (2015). Global ansc ip ome analysis e eals small RNAs
a ec ing Neisse ia meningi idis bac e emia. PLoS ONE 10:e0126325.
doi: 10.1371/jou nal.pone.0126325
Heidenblad, M., Lindg en, D., Jonson, T., Liedbe g, F., Vee la, S., Chebil, G.,
e al. (2008). Tiling esolu ion a ay CGH and high densi y exp ession
p o iling o u o helial ca cinomas delinea e genomic amplicons and candida e
a ge genes speci ic o ad anced umo s. BMC Med. Genomics 1:3.
doi: 10.1186/1755-8794-1-3
Hube , W., Toedling, J., and S einme z, L. M. (2006). T ansc ip mapping
wi h high-densi y oligonucleo ide iling a ays. Bioin o ma ics 22, 1963–1970.
doi: 10.1093/bioin o ma ics/b l289
Johnson, W. E., Li, W., Meye , C. A., Go a do, R., Ca oll, J. S., B own, M., e al.
(2006). Model-based analysis o iling-a ays o ChIP-chip. P oc. Na l. Acad.
Sci. U.S.A. 103, 12457–12462. doi: 10.1073/pnas.0601180103
Kampa, D., Cheng, J., Kap ano , P., Yamanaka, M., B ubake , S., Cawley,
S., e al. (2004). No el RNAs iden i ied om an in-dep h analysis o he
ansc ip ome o human ch omosomes 21 and 22. Genome Res. 14, 331–342.
doi: 10.1101/g .2094104
Land, M., Hause , L., Jun, S. R., Nookaew, I., Leuze, M. R., Ahn, T. H., e al. (2015).
Insigh s om 20 yea s o bac e ial genome sequencing. Func . In eg . Genomics
15, 141–161. doi: 10.1007/s10142-015-0433-4
Liu, X. S. (2007). Ge ing s a ed in iling mic oa ay analysis. PLoS Compu . Biol.
3:30183. doi: 10.1371/jou nal.pcbi.0030183
Loman, N. J., and Pallen, M. J. (2015). Twen y yea s o bac e ial genome
sequencing. Na . Re . 13, 787–794. doi: 10.1038/n mic o3565
Mellin, J., McClu e, R., Lopez, R., G een, D., Reinha d, O. B., and Genco, C.
(2010). Role o H q in i on-dependen and -independen gene egula ion
in Neisse ia meningi idis.Mic obiology 156, 2316–2326. doi: 10.1099/mic.0.03
9040-0
Royce, T. E., Ca ie o, N. J., and Ge s ein, M. B. (2007). An e icien
pseudomedian il e o iling mic o ays. BMC Bioin o ma ics 8:186.
doi: 10.1186/1471-2105-8-186
Selinge , D. W., Cheung, K. J., Mei, R., Johansson, E. M., Richmond, C. S.,
Bla ne , F. R., e al. (2000). RNA exp ession analysis using a 30 base pai
esolu ion Esche ichia coli genome a ay. Na . Bio echnol. 18, 1262–1268.
doi: 10.1038/82367
Smy h, G. K., and Speed, T. (2003). No maliza ion o cDNA mic oa ay da a.
Me hods 31, 265–273. doi: 10.1016/S1046-2023(03)00155-5
Thomassen, G. O., Rowe, A. D., Lagesen, K., Lind all, J. M., and Rognes, T. (2009).
Cus om design and analysis o high-densi y oligonucleo ide bac e ial iling
mic oa ays. PLoS ONE 4:e5943. doi: 10.1371/jou nal.pone.0005943
Tushe , V. G., Tibshi ani, R., and Chu, G. (2001). Signi icance analysis o
mic oa ays applied o he ionizing adia ion esponse. P oc. Na l. Acad. Sci.
U.S.A. 98, 5116–5121. doi: 10.1073/pnas.091062498
Webe , M., Da ies, J. J., Wi ig, D., Oakeley, E. J., Haase, M., Lam, W. L., e al.
(2005). Ch omosome-wide and p omo e -speci ic analyses iden i y si es o
di e en ial DNA me hyla ion in no mal and ans o med human cells. Na .
Gene . 37, 853–862. doi: 10.1038/ng1598
Zhang, Z. H., Jha e i, D. J., Ma shall, V. M., Baue , D. C., Edson, J., Na ayanan, R.
K., e al. (2014). A compa a i e s udy o echniques o di e en ial exp ession
analysis on RNA-Seq da a. PLoS ONE 9:e103207. doi: 10.1101/005611
Con lic o In e es S a emen : The au ho s decla e ha he esea ch was
conduc ed in he absence o any comme cial o inancial ela ionships ha could
be cons ued as a po en ial con lic o in e es .
Copy igh © 2016 Bo ini, Del To dello, Fagnocchi, Dona i and Muzzi. 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 (CC BY). The use, dis ibu ion o ep oduc ion in o he o ums is pe mi ed,
p o ided he o iginal au ho (s) o licenso a e c edi ed and ha he o iginal
publica ion in his jou nal is ci ed, in acco dance wi h accep ed academic p ac ice.
No use, dis ibu ion o ep oduc ion is pe mi ed which does no comply wi h hese
e ms.
F on ie s in Molecula Biosciences | www. on ie sin.o g 7Decembe 2016 | Volume 3 | A icle 82
