HBAT 2: A Py hon Package o analyse Hyd ogen Bonds and
O he Non-co alen In e ac ions in Mac omolecula
S uc u es
Abhishek Tiwa i1*
1Independen Resea che
* abhishek@abhishek- iwa i.com
Abs ac
Hyd ogen bonds and o he non-co alen in e ac ions play a c ucial ole in main aining
he s uc u al in eg i y and unc ionali y o biological mac omolecules such as p o eins
and nucleic acids. Accu a e iden i ica ion and analysis o hyd ogen bonds and o he
non-co alen in e ac ions a e essen ial o unde s anding molecula in e ac ions, p o ein
olding, and d ug design. HBAT (Hyd ogen Bond Analysis Tool) is a widely used
so wa e o analysing hyd ogen bonds and o he weak in e ac ions in mac omolecula
s uc u es. In his pape , we p esen HBAT 2, an upda ed Py hon eimplemen a ion o
he o iginal HBAT ool.
DOI
10.5281/zenodo.17645321
Published
2025-11-18
So wa e
•Code Reposi o y
•Documen a ion
•PyPI Package
1 Summa y 1
HBAT 2 is a Py hon package o au oma ed analysis o hyd ogen bonds and o he 2
non-co alen in e ac ions in mac omolecula s uc u es, a ailable in P o ein Da a Bank
3
(PDB) ile o ma . O iginally de eloped in Pe l/Tk and published in 2007 [1], HBAT 2
4
has been comple ely ew i en in Py hon. The so wa e iden i ies and analyses 5
adi ional hyd ogen bonds, weak hyd ogen bonds, halogen bonds, X-H· · ·π, and π-π6
s acking, and n→π* in e ac ions using geome ic c i e ia. I also de ec s coope a i i y 7
and an icoope a i i y chains and ende s hem as 2D isualisa ions. The la es e sion
8
o HBAT o e s imp o ed c oss-pla o m
kin e
-based g aphical use in e ace (GUI), a
9
simple command-line in e ace (CLI), and a de elope - iendly API making i accessible
10
o use s wi h di e en compu a ional backg ounds and needs. 11
2 S a emen o need 12
Hyd ogen bonds and o he non-co alen in e ac ions a e undamen al o p o ein 13
s uc u e, s abili y, and unc ion. Wi h o e 200,000 s uc u es in he P o ein Da a 14
Bank [2], he e is an inc easing need o au oma ed ools o analyse hese in e ac ions 15
sys ema ically. 16
The landscape o hyd ogen bond analysis ools is di e se bu agmen ed. Classic 17
ools like HBPLUS [3] and HBexplo e [4] pionee ed au oma ed H-bond de ec ion bu 18
lack mode n in e aces and suppo o a di e se ange o in e ac ions. Mo e ecen ools
19
se e specialized niches: PLIP [5] and A peggio [6] excel a p o ein-ligand in e ac ions 20
bu a e web-based wi hou s andalone GUI op ions; HBonanza [7], HBCalcula o [8],21
1/6
and BRIDGE2 [9] ocus on molecula dynamics ajec o ies a he han s a ic 22
s uc u es; MDAnalysis [10], GROMACS [11], and AMBER [12] p o ide H-bond 23
analysis wi hin la ge MD sui es. Tools like VMD [13] and Chime aX [14] o e 24
in e ac i e hyd ogen bond isualisa ion bu limi ed s a is ical analysis. P o einTools 25
[15] p o ides web-based ne wo k analysis bu lacks c oss-pla o m desk op capabili ies. 26
Despi e his ecosys em, he e emains a gap o a comp ehensi e and eliable 27
c oss-pla o m desk op ool ha : (1) analyses di e se in e ac ion ypes beyond 28
canonical hyd ogen bonds, (2) p o ides bo h g aphical and command-line in e aces o
29
di e en use wo k lows, (3) iden i ies po en ial coope a i i y, (4) in eg a es seamlessly
30
wi h he exis ing scien i ic ecosys em, and (5) suppo s lexible pa ame e cus omiza ion
31
wi h domain-speci ic p ese s. 32
HBAT 2 add esses hese limi a ions by p o iding a mode n, Py hon implemen a ion
33
ha in eg a es seamlessly wi h con empo a y s uc u al biology wo k lows. The 34
so wa e is pa icula ly aluable o esea che s in s uc u al biology, compu a ional 35
chemis y, and d ug design who need de ailed analysis o molecula in e ac ions. 36
The o iginal HBAT [1] was de eloped in Pe l/Tk wi h a Windows-only GUI, limi ing
37
i s adop ion in mode n compu a ional en i onmen s. This upda e add esses his 38
limi a ion by making HBAT 2 c oss-pla o m wi h suppo o Windows, Linux, and 39
Mac. 40
Figu e 1. The la es upda e o HBAT 2 uses kin e o p o ide a c oss-pla o m g aphical use in e ace (GUI)
2/6
3 Key Enhancemen s 41
HBAT 2 in oduces se e al key imp o emen s o e he o iginal 2007 e sion: 42
3.1 S uc u e P epa a ion 43
HBAT 2 uses PDBFixe [16],[17] and OpenBabel [18] o au oma ically enhance 44
mac omolecula s uc u es by adding missing a oms, con e ing esidues, and cleaning 45
up s uc u al issues. These capabili ies a e pa icula ly aluable when wo king wi h 46
c ys al s uc u es missing hyd ogen a oms, low- esolu ion s uc u es wi h incomple e 47
side chains, s uc u es con aining non-s anda d amino acid esidues, and s uc u es wi h
48
unwan ed ligands o con aminan s. This in eg a ion imp o es he quali y o analysis. 49
3.2 In e ac ion Co e age 50
HBAT 2 analyses a b oad spec um o in e ac ions including hyd ogen bonds (O-H
· · ·
O,
51
N-H· · ·O, N-H· · ·N, C-H···O), halogen bonds (C-X···Y whe e X=F,Cl,B ,I), X-H···π52
in e ac ions wi h a oma ic sys ems, π-πs acking [19],[20], ca bonyl-ca bonyl n→π*53
in e ac ions [21],[22], and n→π* in e ac ions [23]. This comp ehensi e app oach 54
add esses he g owing ecogni ion o weak in e ac ions’ impo ance in p o ein s uc u e
55
and s abili y [24],[25],[26].56
3.3 Coope a i i y Chains 57
HBAT 2 o e s wo ways o isualise hyd ogen bond ne wo ks: Ne wo kX 58
[27]/Ma plo lib [28] and G aphViz [29]. Unlike ools ha p o ide only basic 59
isualisa ion (VMD, Chime aX) o ocus solely on MD ajec o y dynamics (BRIDGE2,
60
HBonanza), HBAT 2 emphasises coope a i i y chains and ne wo k opology in s a ic 61
s uc u es wi h cus omizable layou s and high- esolu ion expo (PNG, SVG, PDF). 62
3.4 Pa ame e P ese s 63
Buil -in pa ame e se s op imised o di e en expe imen al condi ions (high- esolu ion
64
X- ay, NMR, memb ane p o eins, d ug design) add ess a common challenge in hyd ogen
65
bond analysis. While o he ools equi e manual pa ame e speci ica ion, HBAT’s p ese
66
sys em makes i accessible o expe imen al s uc u al biologis s while main aining 67
lexibili y o compu a ional expe s. 68
3.5 Flexible Ou pu 69
Mul iple expo o ma s ( ex , CSV, JSON) enable in eg a ion wi h downs eam 70
analysis pipelines and s a is ical so wa e. Combined wi h op ional PDBFixe and 71
OpenBabel in eg a ion o au oma ed hyd ogen addi ion, HBAT 2 p o ides a comple e
72
wo k low om aw PDB iles o publica ion- eady analyses. 73
4 Implemen a ion 74
HBAT 2 employs a modula a chi ec u e wi h sepa a e componen s o PDB pa sing, 75
geome ic analysis, s a is ical compu a ion, and isualisa ion. The co e analysis engine
76
uses e icien nea es -neighbo sea ching wi h con igu able dis ance cu o s, ollowed by
77
geome ic il e ing based on dis ance and angula c i e ia. 78
The so wa e implemen s he same undamen al geome ic app oach as he o iginal 79
e sion [1] bu wi h op imised algo i hms and imp o ed handling. 80
3/6
Figu e 2. An example isualisa ion o po en ial coope a i i y chain gene a ed by HBAT 2 so wa e o P o ein Da a
Bank (PDB) en y 6RSA
5 Impac and Applica ions 81
Since i s o iginal publica ion, HBAT has been ci ed in nume ous s udies o p o ein 82
s uc u e and molecula ecogni ion [1]. The la es upda e, howe e , ex ends his 83
impac by p o iding mode n ools o s uc u e-based d ug design, p o ein enginee ing,
84
molecula dynamics analysis, c ys allog aphic s udies, and compa a i e s uc u al 85
analysis. 86
The so wa e’s p ese con igu a ions and lexible pa ame e sys em make i accessible
87
o bo h compu a ional expe s and expe imen al s uc u al biologis s, b oadening i s 88
po en ial use base compa ed o he o iginal e sion. 89
6 A ailabili y 90
HBAT 2 is eely a ailable o download om Gi Hub and PyPI unde he MIT license 91
wi h de ailed use and API documen a ion. The so wa e can be ins alled ia PyPI 92
(
pip ins all hba
) o ia Conda (
conda ins all −c hba hba
), wi h op ional G aphViz
93
in eg a ion o ad anced isualiza ion ea u es. 94
7 Acknowledgemen s 95
The au ho hanks he o iginal co-de elope Sunil K. Panig ahi and acknowledges he 96
s uc u al biology communi y o eedback ha guided he mode niza ion o HBAT. 97
4/6
Re e ences 98
[1]
A. Tiwa i and S. K. Panig ahi, “HBAT: A comple e package o analysing s ong
and weak hyd ogen bonds in mac omolecula c ys al s uc u es,” In Silico Biology,
ol. 7, no. 6, pp. 651–661, 2007, doi: 10.3233/ISI-2007-00337.
99
[2]
H. M. Be man e al., “The p o ein da a bank,” Nucleic acids esea ch, ol. 28,
no. 1, pp. 235–242, 2000.
100
[3] I. K. McDonald and J. M. Tho n on, “Sa is ying hyd ogen bonding po en ial in
p o eins,” Jou nal o molecula biology, ol. 238, no. 5, pp. 777–793, 1994.
101
[4]
K. Lindaue , C. Bendic, and J. S¨uhnel, “HBexplo e–a new ool o iden i ying and
analysing hyd ogen bonding pa e ns in biological mac omolecules,” Compu e
applica ions in he biosciences: CABIOS, ol. 12, no. 4, pp. 281–289, 1996.
102
[5]
S. Salen in, S. Sch eibe , V. J. Haup , M. F. Adasme, and M. Sch oede , “PLIP:
Fully au oma ed p o ein–ligand in e ac ion p o ile ,” Nucleic Acids Resea ch, ol.
43, no. W1, pp. W443–W447, Jul. 2015, doi: 10.1093/na /gk 315.
103
[6]
H. C. Jubb, A. P. Higue uelo, B. Ochoa-Mon a˜no, W. R. Pi , D. B. Asche ,
and T. L. Blundell, “A peggio: A Web Se e o Calcula ing and Visualising
In e a omic In e ac ions in P o ein S uc u es,” Jou nal o Molecula Biology,
ol. 429, no. 3, pp. 365–371, Feb. 2017, doi: 10.1016/j.jmb.2016.12.004.
104
[7]
J. D. Du an and J. A. McCammon, “HBonanza: A Compu e Algo-
i hm o Molecula -Dynamics-T ajec o y Hyd ogen-Bond Analysis,” Jou nal
o molecula g aphics & modelling, ol. 31, pp. 5–9, No . 2011, doi:
10.1016/j.jmgm.2011.07.008.
105
[8]
Y. Wang and Y. Wang, “HBCalcula o : A Tool o Hyd ogen Bond Dis ibu-
ion Calcula ions in Molecula Dynamics Simula ions,” Jou nal o Chemical
In o ma ion and Modeling, ol. 64, no. 6, pp. 1772–1777, Ma . 2024, doi:
10.1021/acs.jcim.4c00054.
106
[9]
M. Sieme s and A.-N. Bonda , “In e ac i e In e ace o G aph-Based Analyses
o Dynamic H-Bond Ne wo ks: Applica ion o Spike P o ein S,” Jou nal o
Chemical In o ma ion and Modeling, ol. 61, no. 6, pp. 2998–3014, Jun. 2021,
doi: 10.1021/acs.jcim.1c00306.
107
[10]
“4.3.1. Hyd ogen Bond Analysis — MDAnaly-
sis.analysis.hyd ogenbonds.hbond analysis — MDAnalysis 1.1.0 documen a ion.”
Accessed: Aug. 08, 2025. [Online]. A ailable:
h ps://docs.mdanalysis.o g/
1.1.0/documen a ion_pages/analysis/hyd ogenbonds.h ml
108
[11]
“Gmx hbond - GROMACS 2025.2 documen a ion.” Accessed: Aug. 08, 2025.
[Online]. A ailable:
h ps://manual.g omacs.o g/cu en /onlinehelp/
gmx-hbond.h ml
109
[12]
“Hbond – AMBER.” May 2020. Accessed: Aug. 08, 2025. [Online]. A ailable:
h ps://ambe hub.chpc.u ah.edu/hbond/
110
[13]
“VMD HBonds Plugin, Ve sion 1.2.” Accessed: Aug. 08, 2025. [Online]. A ailable:
h ps://www.ks.uiuc.edu/Resea ch/ md/plugins/hbonds/
111
[14]
“Tool: H-Bonds.” Accessed: Aug. 08, 2025. [Online]. A ailable:
h ps://www.
cgl.ucs .edu/chime ax/docs/use / ools/hbonds.h ml
112
[15]
N. Fe uz, S. Schmid , and B. H¨ocke , “P o einTools: A oolki o analyze p o ein
s uc u es,” Nucleic Acids Resea ch, ol. 49, no. W1, pp. W559–W566, Jul. 2021,
doi: 10.1093/na /gkab375.
113
[16]
“Openmm/pdb ixe .” OpenMM, Aug. 2025. Accessed: Aug. 08, 2025. [Online].
A ailable: h ps://gi hub.com/openmm/pdb ixe
114
5/6
[17]
P. Eas man e al., “OpenMM 4: A Reusable, Ex ensible, Ha dwa e Independen
Lib a y o High Pe o mance Molecula Simula ion,” Jou nal o Chemical Theo y
and Compu a ion, ol. 9, no. 1, pp. 461–469, Jan. 2013, doi: 10.1021/c 300857j.
115
[18]
N. M. O’Boyle, C. Mo ley, and G. R. Hu chison, “Pybel: A Py hon w appe o
he OpenBabel chemin o ma ics oolki ,” Chemis y Cen al Jou nal, ol. 2, no.
1, p. 5, Dec. 2008, doi: 10.1186/1752-153X-2-5.
116
[19]
G. B. McGaughey, M. Gagn´e, and A. K. Rapp´e, “
π
-S acking In e ac ions: ALIVE
AND WELL IN PROTEINS*,” Jou nal o Biological Chemis y, ol. 273, no. 25,
pp. 15458–15463, Jun. 1998, doi: 10.1074/jbc.273.25.15458.
117
[20]
R. M. Ve non e al., “Pi-Pi con ac s a e an o e looked p o ein ea u e ele an o
phase sepa a ion,” eLi e, ol. 7, p. e31486, doi: 10.7554/eLi e.31486.
118
[21]
A. Rahim, P. Saha, K. K. Jha, N. Sukuma , and B. K. Sa ma, “Recip ocal
ca bonyl–ca bonyl in e ac ions in small molecules and p o eins,” Na u e Commu-
nica ions, ol. 8, no. 1, p. 78, Jul. 2017, doi: 10.1038/s41467-017-00081-x.
119
[22]
R. W. Newbe y and R. T. Raines, “The n→
π
* In e ac ion,” Accoun s o
Chemical Resea ch, ol. 50, no. 8, pp. 1838–1846, Aug. 2017, doi:
10.1021/acs.accoun s.7b00121.
120
[23]
A. Choudha y, D. Gandla, G. R. K ow, and R. T. Raines, “Na u e o Amide
Ca bonyl-Ca bonyl In e ac ions in P o eins,” Jou nal o he Ame ican Chemical
Socie y, ol. 131, no. 21, pp. 7244–7246, Jun. 2009, doi: 10.1021/ja901188y.
121
[24]
G. Desi aju and T. S eine , The Weak Hyd ogen Bond. Ox o d Uni e si y P ess,
2001. doi: 10.1093/acp o :oso/9780198509707.001.0001.
122
[25]
G. Ca allo e al., “The Halogen Bond,” Chemical Re iews, ol. 116, no. 4, pp.
2478–2601, Feb. 2016, doi: 10.1021/acs.chem e .5b00484.
123
[26]
M. B andl, M. S. Weiss, A. Jabs, J. S¨uhnel, and R. Hilgen eld, “C-h?
π
-in e ac ions
in p o eins,” Jou nal o Molecula Biology, ol. 307, no. 1, pp. 357–377, Ma .
2001, doi: 10.1006/jmbi.2000.4473.
124
[27]
A. Hagbe g, P. Swa , and D. S. Chul , “Explo ing ne wo k s uc u e, dynamics,
and unc ion using Ne wo kX,” in P oceedings o he 7 h py hon in science
con e ence, Pasadena, CA USA, 2008, pp. 11–15.
125
[28]
J. D. Hun e , “Ma plo lib: A 2D g aphics en i onmen ,” Compu ing in science &
enginee ing, ol. 9, no. 3, pp. 90–95, 2007, doi: 10.1109/MCSE.2007.55.
126
[29]
G aph iz Con ibu o s, “G aph iz - g aph isualiza ion so wa e.”
h ps://
g aph iz.o g/, 2024.
127
6/6
