Synthetic, spectral, antimicrobial and QSAR studies on novel mannich bases of glutarimides

Article abstract of DOI:10.4067/S0717-97072010000300001

A series of forty Mannich bases of glutarimides with sulfonamides and secondary amines were synthesized and evaluated in vitro against six pathogenic Gram-positive and Gram-negative bacteria. The synthesized Mannich bases were characterized by elemental a

Full text of DOI:10.4067/S0717-97072010000300001

J. Chil. Chem. Soc., 55, Nº 3 (2010)
SYNTHETIC, SPECTRAL, ANTIMICROBIAL AND QSAR STUDIES ON NOVEL
MANNICH BASES OF GLUTARIMIDES
ANJU DAS MANIKPURIa, SHEELA JOSHIb AND PADMAKAR V. KHADIKAR*c
aDepartment of chemistry, ISLE,IPS Academy,Knowledge city,Rajendranagar, Indore-452017,MP,India.E.mail:anju.dm@rediffmail.com
bSchool of Chemical Sciences, Devi Ahilya University, Takshashila Campus, Khandwa Road, Indore-452017,MP, India.E.mail: spjoshi11@rediffmail.com
cLaxmi fumigation and pest control centre,Research Division,3 Khatipura, Indore- 452017, MP,India. E.mail:pvkhadikar@rediffmail.com.
(Received: July 28, 2009 - Accepted: July 20, 2010)
ABSTRACT
A series of forty Mannich bases of glutarimides with sulfonamides and secondary amines were synthesized and evaluated in vitro against six pathogenic Gram-
positive and Gram-negative bacteria. The synthesized Mannich bases were characterized by elemental and spectral analysis. The modeling anti-bacterial activities
1
of these newly synthesized Mannich bases against six bacteria was attempted employing H NMR chemical shift, physicochemical properties and topological
indices as the correlating parameters. Our results, based on Quantitative Structure-Activity Relationships (QSARs), have indicated that statistically significant
models are obtained for modeling the anti-bacterial activities The results are discussed critically using a variety of statistical parameters.
Keywords: Glutarimides, sulfonamides, Mannich bases, antimicrobial activity, QSAR.
13C NMR chemical shift as a molecular descriptor and secondly to use the
same for modeling property-activity- toxicity of organic compounds acting
INTRODUCTION
as drugs. One of such applications studied by Khadikar being modeling CA
inhibition using 13C NMR chemical shift 6. Prompted by these results we
have undertaken the present study, in that we have investigated variance of
antibacterial activity using ∑NMR chemical shifts as one of the correlating
parameters. We have also observed that in many cases ∑NMR chemical shifts
in combination with physicochemical parameters as well as topological indices
improved results are obtained. In doing so we have used maximum -R2 method
and applied variety of statistics10. We have, therefore, attempted modeling of
antibacterial activities of the newly synthesized Mannich bases using ∑NMR
chemical shifts, physicochemical parameters as well as topological indices as
the correlating parameters. At this stage it is interesting to mention that some
people cretised the use of topological indices and consider their use in the
development of the models as quite restrictive, since these descriptors only
depend on 2D molecular features and not in conformational or electronic
properties that are most expected to govern the biological activity of the
compounds under study than topological features. Such a criticism is due
to the fact that generally no physical significance is attached to topological
indices. However, there are several cases in that topological indices correlate
excellently with conformational, electronic and other related properties making
their judies use in QSAR.
The Quantitative Structure-Activity Relationships (QSARs) are
mathematical models relating measured biological activity of series of
structurally related compounds / pharmacological agents to the variation in
their chemical structure. In cases in which some physico-chemical properties
or toxicities of such compounds are related to their structures, the methodology
is called Quantitative Structure-Property relationships (QSPRs) or Quantitative
Structure-Toxicity Relationships (QSTRs), respectively. Such methodologies
are widely used in environmental toxicology to understand the adverse effects
of chemical compounds. The QSPR /QSAR /QSTR methodology is very useful
for a large number of untested chemicals present in nature, and because of the
high costs of biological testing 1-19. QSAR models are nowadays regarded as a
scientifically credible tool for predicting and classifying biological activities of
untested chemicals. QSAR has become inexorably embedded as an essential
tool in the pharmaceutical industry, from lead discovery, optimization to lead
development and computer –aided drug designing 20, 21. A growing trend is to
use QSAR early in the drug discovery process as a screening and enrichment
tool to estimate from further development those chemicals lacking drug
like properties 21 or those chemicals predicted to elicit a toxic response. The
fundamental assumption of QSAR is that variations in the biological activity of
a series of chemicals that target a common mechanism of action are correlated
with variations in their structural, physical and chemical properties 22. These
biological activity of structurally related chemical compound in terms of in-vivo
or in-vitro properties can be determined by experimental or more efficiently
by computational mean.. Needless to state that a statistically validated QSAR
model is capable of predicting the biological activity of a new chemical within
the same series in lieu of the time-consuming and lab our-intensive processes
of chemical synthesis and biological evaluation. Applied judiciously, QSAR
can save substantial amount of time, money, and human resources.
The increasing popularity of the Mannich reaction as well as utility of
Mannich bases has been fluted by the ubiquitous nature of nitrogen in drugs and
natural products as well as by the potential of this multi-component reaction to
generate diversity. Interest in Mannich bases has been quite attractive and wide
ranged considering the enormous domain of the applications involving variant
biological 31-35, pharmaceutical 36-40 and industrial41-43 properties. Glutarimides
moiety with the intact imide group is acting as the carrier molecule (vector),
which transports biologically active substitutents (functional groups) through
cell membranes44. It is also a component of newly synthesized antibiotics, which
exert antiviral and antifungal activity 45, 46. Glutarimides (2,6-piperidinedione)
moiety is found in a number of antibiotics with antiviral and fungicidal
The molecular structure and NMR chemical shift information of organic
compounds acting as drugs can be combined to form powerful models
of biological activity. Such data-activity relationship is now-a-day called
activity47-50
. Furthermore, the 2,6-piperidinedione moieties constitute an
Quantitative Structure-Data-Activity Relationship (QSDARs) in place of
23-30
important centre in several new anticancer drugs which have recently been
introduced into experimental chemotherapy51-53. It is also a structural part of
a number of molecules with interesting biochemical activity54. In view this
we have attempted aminoalkyaltion of various glutarimides-moieties with
sulphonamides and secondary amines. The structural characterization these
QSAR as it involved the use of spectroscopic data. As is well known
,
chemical shifts in NMR offer a powerful probe for the study of the immediate
atomic environment in a molecule. It is worthy to mention that NMR spectra
reflect quantum mechanical properties and that QSAR depends on local
electrostatics and geometry of the molecule. The 13C NMR spectrum of a
compound contains a pattern of frequencies that correspond directly to the
quantum mechanical properties of the carbon nuclear magnetic dipole in a
magnetic field. The spectral pattern reflects the local electrostatic environment
and electron orbital configuration of each atom. The resonance from different
carbon orbital configurations is generally well-separated from each other,
which permits the use of advantageous for 13C NMR spectral directly to build
1
newly synthesized Mannich bases is made using elemental, UV, IR and H
NMR studies.
The anti-bacterial potential vis-à-vis QSAR study of the newly synthesized
Mannich bases in the present is made in the following four different ways:
(i ) QSAR study based on using ∑NMR chemical shifts as molecular
descriptor;
the QSDAR models23, 24
.
(ii) QSAR study based on using physicochemical parameters as molecular
descriptor;
(iii) QSAR study based on using topological indices as molecular
Recently one of the authors (PVK) has initiated interesting investigations
on 13C NMR chemical shift 25 - 30. His approach was two-fold: firstly to establish
e-mail:pvk3615@rediffmail.com
283

J. Chil. Chem. Soc., 55, Nº 3 (2010)
descriptor, and (iv) QSAR study based on combination of ∑NMR chemical
shifts, with physicochemical properties and or topological indices.
The results obtained are presented in Tables 1-7.
EXPERIMENTAL SECTION
General information.
All the m.p. of the synthesized Mannich bases ware determined using
Thomas Hoover capillary melting point apparatus and their purity was
ascertained by TLC method. The antimicrobial screening was performed using
paper disc method and the results were statistically evaluated using variety of
statistical parameters. Mullar Hinton Agar was taken as media for cultivation
of bacteria. The inhibitory effect of the samples were measured against each of
the bacteria after incubation for 24 hours at 370C.The experiments were run in
triplicate and the mean of readings were recorded.
Synthesis of Mannich bases from primary amines.
Mannich bases of glutarimides were prepared by reacting various
glutarimides (0.01mol) dissolved in 20 mL of ethanol with sulfonamide
(0.01mol). 2.5 mL (0.01mol) of formaldehyde solution (37%, v/v) was added
slowly with constant stirring. The pH of the mixture was adjusted to 3.5 by
adding 0.5 mL of 1 mol L-1 HCl. The mixture was kept at efficient ice cooling
for half an hour, and then refluxed on water bath. Reflux time varied with the
sulfonamide used. The refluxed mixture was kept at 00C for four days when
crystalline product was obtained. The product was re-crystallized from dry
distilled ethanol and dioxane-water (1:1). The compounds were characterized
by elemental, UV, IR and NMR studies.
Scheme 2: Synthesis of Mannich bases from secondary amines.
Synthesis of Mannich bases from secondary amines.
Secondary amine (0.01mol) was added to an ethanolic solution (50 mL) of
glutarimides ( 0.01 mol) in a flat bottom flask. Amount of 0.4 mL (0.015mol) of
formaldehyde solution (37%, v/v) was added slowly with constant stirring. The
reaction mixture was stirred at 70-750C for 3.0 to 8.5 hours, depending upon the
secondary amine. The remaining portion of formaldehyde solution was added
in two installments after 1 and 2 hours, respectively. The reaction mixture was
kept overnight in the refrigerator. Next day, the excess of solvent was distilled
off from the reaction mixture under reduced pressure. It was again kept for
crystallization in the refrigerator. The products obtained were purified by re-
crystallization from dry distilled ethanol. The compounds were characterized
by elemental, UV, IR and NMR studies. For details see schemes 1 and 2.
Spectral studies.
The 1H NMR spectra in DMSO and CDCl solvent were recorded on
Bruker DRX-300 FT NMR Spectrometer. The U3V spectra were recorded on
Schimadzu UV-160A, UV-visible spectrophotometer.
Antimicrobial activity .
The sum of the NMR chemical shift ,∑NMR, and the anti-bacterial activity
for the set of 40 Mannich bases as given in Table 1.The useful descriptors
(∑NMR, physical parameters and topological indices) from a large set
were chosen using multiple variable selection and are given in Table 2. The
regression analyses were performed starting from a single to ten descriptors
and the regression reports are given as supplementary material. The plots of
number of descriptors against R2 (Fig. 1) indicated that eight variable model is
the most appropriate model for modeling anti-bacterial activity against all the
six bacteria.. This is in accordance with the rule of Thumb 63 which states that
the descriptors to be used in multiple regression analysis should be one-fifth
of the total number of compounds under study. The detail regression analysis
of the models is given in the text under Results and Discussion section. The
correlelation matrices for the parameters involved in these models are also
presented separately as supplementary material. Finally, Ridge parameters for
each of these models are also mention as supplementary material.
Table 1. Antibacterial activity of the Mannich bases and their and NMR-
Chemical shift (1-40).
B.
subtilis
S.
K. pneu- P. auru-
C.N. ∑NMR
S. typhi E.coli
aureus moniae ginosa
1
2
33.98
34.26
31.68
32.5
10.5
10.0
10.5
11.0
12.0
13.0
11.0
10.5
10.0
11.2
11.6
11.8
13.2
14.8
10.5
12.0
10.0
10;.0
11.0
11.5
10.5
8.5
10.0
12.o
11.0
8.5
11.5
11.0
12.0
9.5
10.8
11.5
12.0
13.0
13.5
14.2
15.8
10.8
10.2
11.8
13.0
13.5
12.5
12.0
10.5
10.0
10.5
11.0
12.0
13.0
11.0
10.5
10..0
11.2
11.6
11.8
13.2
14.8
3
4
5
34.29
30.42
36.44
34.36
30.88
32.48
30.98
31.28
36.92
32.22
7.5
10.5
10.0
10.5
10.5
10.0
11.5
12.4
11.0
11.2
11.5
6
10.0
10.5
11.0
11.5
11.5
11.0
10.0
10.5
12.0
Scheme 1: Synthesis of Mannich bases from sulphonamides.
7
8
9
10.0
11.0
11.0
11.0
11.5
12.5
10
11
12
13
14
284

J. Chil. Chem. Soc., 55, Nº 3 (2010)
Codes for the bacteria
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
30.24
34.60
34.28
32.60
34.46
30.92
30.28
30.60
32.42
32.72
27.04
34.42
30.90
34.90
32.48
36.5
15.6
10.5
8.5
13.5
10.0
10.0
10.0
10.0
10.0
11.5
12.5
13.5
12.5
-
12.5
10.5
10.5
10.5
10.0
11.0
11.0
11.5
12.5
11.5
-
12.0
11.0
11.0
10.5
11.0
11.5
11.5
12.0
12.5
11.0
-
12.0
11.5
11.5
11.6
11.5
11.5
11.5
12.0
11.5
11.0
-
15.6
10.5
8.5
Code
Bacteria
C32
C33
C34
C35
C36
C37
B.subtilis
S.typhi
E.coli
S.aureus
K.pneumoniae
P.aeruginosa
7.5
7.5
10.5
12.0
13.4
11.0
15.5
14.8
-
10.5
12.0
13.4
11.0
15.5
14.8
-
12.0
12.5
11.5
11.0
10.0
10.8
10.4
10.6
10.8
10.9
10.6
10.4
10.4
10.0
10.0
13.5
12.5
10.5
10.5
10.0
10.8
13.8
11.8
12.8
13.8
11.8
14.8
13.8
12.8
11.8
12.5
10.0
11.0
10.5
10.5
12.2
11.2
13.2
12.2
13.2
14.2
15.2
11.2
10.2
12.2
12.0
12.5
12.5
11.5
14.5
13.5
12.5
16.5
17.5
12.5
11.5
10.5
13.5
12.5
13.5
12.5
12.0
12.5
12.5
12.5
11.5
10.5
12.0
12.5
12.5
11.5
10.5
13.0
13.5
11.5
12.0
12.5
11.5
11.0
10.0
10.5
15.5
12.5
11.5
11.5
11.5
11.5
11.5
11.5
11.5
30.26
34.62
34.42
30.24
32.4
30.92
30.24
30.2
34.92
34.9
Table 2. Useful variables from the larger pool and their codes.
Code Variable Symbol
Name of the variable
A
B
C
D
E
F
G
H
I
C2
C9
SNMR
d
α
Sum of NMR
Density
Polarizability
Monoisotopic Mass
Nominal Mass
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
C21
C22
C23
C24
C25
C26
C27
MM
NM
AM
ZM1
ZM2
Pol
SMT1
Xu
SP1
W
Average Mass
First Zagreb index M1
Second Zagreb index M2
Polarity number
Schultz Molecular Topological Index
Xu Index
Superpendentic Index
Wiener Index
Harary Index
Zero order randic connectivity Index
First order randic connectivity Index
Second order randic connectivity Index
Third order randic connectivity Index
Fourth order randic connectivity Index
Zero order randic valence connectivity
Index
J
K
L
M
N
O
P
Q
R
S
Har
0c
1c
2c
3c
4c
T
0cv
U
V
W
X
C28
C29
C30
C31
1cv
2cv
3cv
4cv
First order randic valence connectivity
Index
Second order randic valence connectivity
Index
Third order randic valence connectivity
Index
Fourth order randic valence connectivity
Index
285

J. Chil. Chem. Soc., 55, Nº 3 (2010)
topological indices better quality models are obtained and that most appropriate
models are obtained when all the three types of molecular descriptors are used
together. We observed that if ∑NMR chemical shifts are eliminated from the
most appropriate model the quality of the model is decreased significantly. This
exhibits the dominating role of ∑NMR chemical shifts. The proposed models
are given below:
(i) Model for B.subtilis
Anti-bacterial activity(B.subtilis) = 6.7268 + 4.6812*10-2(±3.6965*10-2)
C2
+7.0545*10-3(±1.6369*10-2}C3
-3.0702*10-3(±2.107*10-3)C17
+0.01124 (±8.949774*10-3)C20
+0.24859 (±0.3089584)C24
-6.4740*10-2(±0.1287502)C27
+1.1076*10-4(±1.4978*10-4)C28
+0.33949 (±0.2729939)C30
N=40,CV=0.1080532,R2=0.2601,R2A=0.0000,F=0.7470
(ii) Model for S.typhi
Anti-bacterialactivity(S.typhi)=9.888266-4.510451*10-2 (±2.321028*10-
2) C2
+1.473106(±0.2763391)C16
-1.525129(±0.4197105)C21
+0.2966534(±0.2012577)C24
+2.086771(±1.280614)C25
-0.3372007(±0.1177242) C26
-0.2197946(±6.764524*10-2) C27
+2.066341*10-4(± 9.204068*10-5) C28
N=40,CV= 6.8970,R2=0.8328,R2A=0.7541, F=10.5844
(iii) Model for E.Coli
Anti-bacterial activity(E.coli) = 17.3471-8.5342*10-2(±3.2712*10-2)C2
-8.2224(±4.0231C9
+12.3122 (±8.025793) C11
+12.1831 (±8.015417) C12
-4.3494*10-2(±4.455554*10-2) C15
-1.3177 (±0.5166374)C18
+2.6913*10-4(±1.0817*10-4)C28
-0.3951 (±0.1142 )C29
N=40,CV= 7.9712,R2=0.7836,R2A=0.6817, F=7.6935
(iv) Model for S.aureus
Anti-bacterialactivity(S.aureus)=32.7323+0.1425 (±3.5039*10-2)C2
+1.8960 (±0.2853)C16
-1.0472*10- 2(±2.1726*10-3)C17
+4.9660*10-2(±9.5501*10-3)C20
- 5.5463 (±0.8328)C22
-0.3017 (±0.2427 )C24
-3.7687*10-4(±1.1822*10-4 )C28
-0.5538 (±0.1980) C31
N=40,CV= 7.7740,R2=0.7898,R2A=0.6909, F=7.9863
(v) Model for K.pneumoniae
Anti-bacterialactivity(K.pneumoniae) =6.3773+ 0.7490 (±0.2062) C2
-15.4731 (±7.2948) C10
+15.4189(±7.2829) C11
-5.4125*10-3(±1.9237*10-3) C12
+2.0276*10-2(±8.3083*10-3)C17
Figure 1. Correlation of number of descriptors (variable count) with R2
Supplementary material. (1) Variable selection for multiple regression analysis.
+0.3188 (±0.2348 )C20
-3.1490*10-4(±1.1616*10-4) C28
N=40,CV= 7.1296,R2=0.6142,R2A=0.4326, F=3.3825
RESULTS AND DISCUSSION
(vi) Model for P.aeruginosa
Anti-bacterialactivity(P.aeruginosa) =6.8045+10.9606 (±0.3556)C16
+1.0584 (±0.4314)C18
A perusal of the Table 1 demonstrates the sequence of activity in each of
the bacteria used in the present study. The data presented in this Table 1 show
that degeneracy in activity exists in each of the six bacteria used. Also that
the sequences of activities are very different for the bacteria used. This data,
therefore, do not exhibit any structure-activity relationship for the variation of
antibacterial activity. This has prompted us to compute molecular descriptors
and then to examine structure activity using QSAR methodology.
Preliminary regression analysis has indicated that use of ∑NMR chemical
shifts alone do not give any statistically significant model. However, when
∑NMR chemical shifts are combined with physical properties and also
-1.9102 (±0.6176)C21
+3.8562 (±1.7120)C25
-0.6298 (±0.3388)C26
-0.1285 (±0.1361) C27
+2.0391*10-4(±1.3594*10-4) C28
+0.6882 (±0.4046)C31
N=40, CV= 9.0854*10-2, R2=0.4846,R2A=0.2420, F=1.9978
286

J. Chil. Chem. Soc., 55, Nº 3 (2010)
Physicochemical significances of the proposed models
CONCLUSIONS
It will be interesting to discuss the physicochemical significances of the
proposed models. We observed that invariably all the models contain C2
(∑NMR) as the correlating parameter. It means, therefore, that ∑NMR play
a dominating role in modeling antibacterial activity against all the bacteria.
That is electronic effect is the main parameter for the exhibition of the activity.
In case of models -3 and model-5, in addition to ∑NMR, physicochemical
parameters along with topological indices are involved in modeling the
antibacterial activities. The physicochemical parameters involved indicate
that the size and shape are responsible for the exhibition o the antibacterial
activities against these bacteria. The topological indices involved in these
and other models are mainly connectivity indices. This means that the extent
of connectivity is the responsible for the exhibition of anti bacterial activity
against all the bacteria used.
The newly synthesized Mannich bases appeared to be very potent and
outstanding antibacterial agents with promising activity and found safer.
These novel Mannich bases could be used as useful drug. Our findings will
prove helpful to those who are engrossed in the synthesis of potential Mannich
bases as drugs with minimum side effects and also having comparatively
low cost. Thus, the result presented in this paper is valuable in constructing
pharmacologically imperative heterocyclic as new exotic drugs. Efforts are
continuing to synthesize new amino-methyl derivatives of various active
hydrogen compounds, that the derived compounds may have enhanced
pharmacological activity. Our results also show that the antibacterial activity
of the Mannich bases could be modeled using sum of the NMR chemical shifts,
physicochemical parameters, and topological indices. The combination of these
parameters gives statistically significant models for modeling antibacterial
activity against S.typhi, E.coli, S.aureus, and K.pneumoniae.
It is also worth mentioning that all the proposed models, except model-5
(which has 7 correlating parameters), all other models contain 8 correlating
parameters. Looking to the size of the sample i.e. the number of compounds
used this is accordance with the rule of thumb. Furthermore, the plot of number
of variables used against R2
ACKNOwLEDGEMENTS
( Fig.1) also favors the use of 7 to 8 correlating parameters.
The aforementioned models were further examined employing Ridge
statistics.
The authors wish to express gratitude to the Director, CDRI, Lucknow,
India for recording elemental analysis.
REFERENCES
Ridge statistics.
Application of Ridge statistics provides important statistical parameters
namely variance inflation factors (VIFs) for each of the parameters involved in
the model. The VIF is defined for each variable in the equation, and not for the
equation as a whole, so there should be as many VIFs, as there are correlating
parameters. The VIF is defined as:
1. C. Hansch, A. Leo, In Exploring QSAR: Fundamental and application
in chemistry and biology ,S.R. Heller, Ed. American Chemical Society:
Washington, DC. (1995)
2. M. Karelson, Molecular descriptors in QSAR/QSPR, Wiely-Interscience
(2000)
3. M.V. Diudea, M.S. Florescn, P.V. Khadikar, Molecular topology and its
applications, EFICON Press: Bucarest. (2006)
VIF = 1(1-R2i)
(19)
4. C. Hansch, B.R. Telzer, L. Zhang, Crit. Rev. Toxicol. 25,67- 89. (1995)
5. H. Branes, Mamalian toxicology-property prediction and QSAR
technique,Peter Fisk Associates Whitestable: Kent. (2004)
6. S. Bradbury, Taxicol. Lett. 79,229-237, (1995)
7. P.R. Duchowicz , E.A. Castro, A.A. Toropov, E. Benfenati, Top
Heterocycl. Chem 3, 1-18 , (2006)
8. M.V. Diudea, (Ed) QSPR/QSAR Studies by Molecular Descriptors, Nova
Science, (2000)
9. J. Devillers, A.T. Balaban, (Eds) Topological Indices and Related
Descriptors in QSAR and QSPR, Gordon & Breach Science Pub.,
Amsterdam (The Netherlands), (1999)
Where Ri is the multiple correlation coefficient of the ith independent
variable on all of the other independent variables. In the proposed models, all
these VIFs should be less than 10 indicating that no co- linearity problem exists
in the model.
The VIFs values for the parameters involved in the aforementioned models
are given supplementary material; which shows that each of the models contain
one or more descriptors having VIFs values larger than 10. Therefore, based
on VIFs values a major problem of co- linearity exists for them. However, the
magnitude of condition number indicates the existing of mild co- linearity. In
such cases Randic recommendations 59-60 are used to arrive at the final decision.
10. M.V. Diudea, (Ed) QSPR/QSAR Studies by Molecular Descriptors, Nova
Science( 2000 )
Randic recommendations
59-60
Randic
stated that if a descriptor strongly correlates with another
11. D. Janezic, Nikolic, T.N.Sonja, Graph theoretical matrices in chemistry,
University of Kragujevac, Faculty of Science, Kragujevac,205 ( 2007)
12. Li, Xueliang , Tianjin, I.Gutman, Mathematical Aspects of Randic-
Type Molecular Structure Descriptors Nankai University, University
of Kragujevac, Kragujevac, Serbia University of Kragujevac,
Faculty of Science, Kragujevac ISBN: 86-81829-67-X, 330 ,(2006)
Rouvray , H. Dennis (Editor), Complexity: Introduction and Fundamentals
University of Georgia, Athens, Georgia, USADanail Bonchev (Editor),
Virginia Commonwealth Univ., ichmond, USACRC PressISBN: 0-415-
28791-X, 208, (2003)
descriptor already used in a regression, such a descriptor in most studies
should be discarded. For example χ and χ , 1χ often strongly correlate and
1
2
2
in many structure-property-activity studies χ has been discarded. This is not
theoretically justified and despite the widespread practice should be stopped.
Although two highly correlated descriptors overall depict the same features of
molecular structure, it is important to recognize that even highly interrelated
descriptors differ in some other structural traits. The difference between them
may be relatively small but nevertheless very important for structure-property
regression.
The criteria for inclusion or exclusion of descriptors should not be based
on parallelism between descriptors even if overwhelming, but should be based
on whether the part in which two descriptors disagree is or is not relevant for
the characterization of the property considered .If the part in which the second
descriptor differ from the first, regardless of how small it is, is relevant for the
property under consideration, then the descriptor should be included. Randic 59,
60 further stated that the selection of descriptors to be used in structure-property-
activity studies should not be delegated solely to computers, although statistical
criteria will continue to be useful for preliminary screening of descriptors taken
from a large pool. Often in an automated selection of descriptors, a descriptor
will be discarded because it is highly correlated with another descriptor already
selected. But what is important is not whether two descriptors parallel one
another; i. e. duplicates much of the same structural information, but whether
they are complementary in those parts that are important for structure-property-
activity correlations. Hence, the residual of the correlation between two
descriptors should be examined and kept or discarded depending on how well
it can improve the correlation based on already selected descriptors.
13. R. King, Bruce (Editor), Topology in Chemistry: Discrete Mathematics
of Molecules University of Georgia, Athens, Georgia, USADennis H.
Rouvray (Editor), University of Georgia, Athens, Georgia, USAALBION/
HORWOOD Pub.ISBN: 1898563764, 400 ,( 2002)
14. H.Diudea,
V
Mircea., I.Gutman, Molecular Topology NOVA
SCIENCEISBN, 332 , (2001)
15. R. Todeschini, Handbook of Molecular Descriptors WILEY-VCHISBN,
668, (2000)
16. L.Kier, R.Hall, Molecular Structure description ACADEMIC PRESS
ISBN,286, (1999)
17. A.T.Balaban , (Editor), From Chemical Topology to Three-Dimensional
Geometry PLENUM PRESSISBN, 420 ,( 1997)
18. G.P.Mezey, Shape in Chemistry: An Introduction to Molecular Shape and
Topology JOHN WILEY & SONS, 224 ,(1993)
19. N.Trinajstic, Chemical Graph Theory CRC Press, 322 ,( 1992)
20. H.D.Rouvray , (Editor), Chemical Graph Theory: Introduction and
Fundamentals USATAYLOR & FRANCIS, 300 ,( 1991)
21. R.B.King , (Editor), Graph Theory and Topology in Chemistry, USA
ELSEVIER, 575 ,( 1987)
287

J. Chil. Chem. Soc., 55, Nº 3 (2010)
22. D. Bonchev, Information Theoretic Indices for Characterization of
Chemical Structures USARESEARCH STUDIES PRESS, 249 , 1983
23. A.T. Balaban, (Editor), Chemical Applications of Graph Theory
ACADEMIC PRESS, 389 ,( 1976)
24. P.V.Khadikar, S.Singh, M. Jaiswal, D. Mandoli, Bioorg. Med. Chem.
Lett, 14, 4795,(2004).
25. P.V. Khadikar, N.S. Sapare, S.Kumar, R.N. Patel, Indian J.Chem.. 39A,
386(2000).
26. M. Jaiswal, P. V. Khadikar, J. Indian Chem. Soc., 82, 247-249(2005).
27. Patel, R.N. Kumar,S. Pande, K.B. Khadikar,P.V. Spectro.Chem.Acta.
58A, 2961,(2001).
(i) B.subtilis
(i)Using variables: C2, C9-C13. (ii) Using variables : C2,C14-C31.
(iii)Using variables : C2,C9-C31.
Model
Size
Coded
Variables
R2
1
2
3
4
5
6
0.1718
0.1722
0.2495
0.2535
0.2595
0.2611
A
AC
BDF
BDEF
ABDEF
ABCDEF
28. S.N.Kane, A.K. Sharma, A. Gupta, P.V.Khadikar, S.M.Ali, Bull.Soc.
Chem.Belg. 105,155(1996).
29. S. Joshi, A .D .Manikpuri, D. Khare, J. Indian Chem.Soc. 85, 508-512 ,
(2008) .
30. S. Joshi, A. D. Manikpuri, P. Tiwari, Bioorganic & Medicinal Chemistry
Letters 17, 645-648, (2007).
31. S Joshi, N. Khosla, D. Khare, R. Sharda, Bioorganic & Medicinal
Chemistry Letters 15, 221-226, (2005).
32. S. Joshi, P.Tiwari, N. Khosla, Bioorganic & Medicinal Chemistry 12,
571,(2004).
33. S. Joshi, N. Khosla,D.Khare, Bioorganic & Medicinal Chemistry Letters
13, 3747, (2003).
34. D. Sriram, P.Yogeeswari, S.P.Reddy, Bioorg.Med.Chem.Lett.16, 2113-
2116,(2006).
35. M. Tramontiny, L. Angiolini, Mannich bases: Chemistry and uses, CRC
Press, Boca Raton, (1994).
36. A. B. McKaguue, J.Appl. Chem.Biotechnol. 24, 607, (1974).
37. M. Park, J.Lee, J. Choi ,Bioorganic & Medicinal Chemistry Letters 6,
1297-1302 (1996) .
Model
Size
Coded
Variables
R2
1
2
3
4
5
6
7
8
9
10
0.0630
0.1113
0.2379
0.2727
0.2929
0.3062
0.3189
0.3277
0.3576
0.3863
A
AR
FIR
FIPR
FIJPR
DFIJPR
DFILPQR
DFILOPQR
DEHIJLOQR
DEHIJLOPQR
Model
Size
Coded
Variables
R2
38. D.Michalska, B.Morzyk, D.C.Bieńko, W.Wojciechowski. Med
Hypotheses. 54 472-4 ,(2000) .
39. T.Otani, Y.Minami, H.Yoshida, H.Nishida, T.Nakayama, K. Ohishi,
H.Tsuda, I. Ikeda, J. Jpn.Soc. Cancer Ther. 23, 1560 (1988).
40. T. Sonoda, M. Kobayashi, M. Ubukata, H. Osada, K. Isono, J. Antibiot.
45, 1963, (1992).
41. T. Otani, Y. Minami, H. Matsumoto, T. Marunaka, Z.-X. Lou and Q.-W.
Yu, J. Antibiot. 42, 654, (1989).
42. K. Sugawara, Y. Nishiyama, S. Toda, N. Komiyama, M. Hatori, T.
Monyama, Y. Sawada, H. Kamei, M. Konishi and T. Oki, J. Antibiot, 45,
1433, (1992).
43. T. Sonoda, H. Osada, J. Uzawa and K. Isono, J. Antibiot. 44, 160, (1991).
44. T. Sonoda, K. Kobayashi, M. Ubukata, H. Osada and K. Isono, J.
Antibiot., 45, 1963,(1992).
45. S.R. Burzynski ,T.T. Hai, Drugs Fut. 10, 103, (1985).
46. S.R. Burzynski, Adv. Exp. C h . Chemother. 6, 45 ,(1988).
47. D.C.Bienko, D. Michalska, S.Roszak, W. Wojciechowski, J.Phys.Chem.
A101, 7834-7841,(1997)
1
2
3
4
5
6
7
8
9
10
0.0693
0.0993
0.1812
0.2136
0.2241
0.2433
0.2623
0.2748
0.2876
0.2896
B
BJ
BJW
ABJW
ABJUW
AJMQUW
ABJMUVW
ABJMQTUW
ABJMQTUVW
ABJMQSTUVW
(ii) S.typhi
(i)Using variables: C2, C9-C13(ii) Using variables: C2, C14-C31.
(iii) Using variables: C2, C9-C31.
Model
Size
Coded
Variables
R2
48. W.B.Taylor, Biometrics 13, 1-12 ,(1957).
49. S. Joshi, N. Khosla, Indian drugs ,31, 548-550,(1994).
50. N. Siddiqui, S. N. Pandeya, S. A. Khan, J. Stable, A. Ran, M. Alam, Md.
Arshad, M. A. Bhat, Bioorg. Med. Chem. Letters, 17,255-259,(2007).
51. S. Joshi, N. Khosla, D. Khare, P.Tiwari, Acta Pharm., 52,197-206,(2002).
52. S. Joshi, N. Khosla, P. Tiwari, Bioorg. Med. Chem., 12,571-576,(2002).
53. ACD-Lab software for calculating the referred Physicochemical
ParametersChem Sketch 3.0,www.acdlabs.com
54. H.Wiener, J. Am. Chem .Soc., 69.17,(1947).
55. L. B. Kier, L.H. Hall, Molecular Connectivity in Chemistry and Drug
Research, Academic: New York, (1976).
56. M.V. Diudea, M.S. Florescu, P.V. Khadikar, Molecular Topology and
Its Applications, EFICON, Bucharest, 412 ,(2006).
57. M. Randic, J. Am.Chem. Soc, 97,6609,(1975).
58. The DRAGON Plus v 55 program is available from Todeschini, R Talete
srl, via Pisani, 13-20124 Milano, Italy
59. HyperChem Software, Hypercube Inc Florida Science and Technology
Park ,1115 NW 4th street Gainesvilla Florida ,326001, USA (2006).
60. S.Chaterjee, A.S.Hadi, B.Price, Regression Analysis by Examples, 3rdEd
Wiley: New York, (2000).
1
2
3
4
5
6
0.2866
0.4293
0.4508
0.4549
0.4592
0.4592
C
EF
AEF
ABEF
ABDEF
ABCDEF
Model
Size
Coded
Variables
R2
1
2
0.3687
I
0.5385
0.6190
0.6797
0.7241
0.7823
0.8114
0.8328
0.8493
0.8584
IN
3
INU
4
AINX
5
AFINU
6
INRSTU
AINRSTU
AINQRSTU
AINQRSTUX
ABINQRSTUX
7
8
61. M. S. Tute, History and Objectives of Quantitative Drug Design in
Advances in Drug Research, N. J. Harter, A. B. Simmord, (Eds),
Academic Press: London 6, 1, (1971)
9
10
62. S. Joshi, A. D. Manikpuri, D. Khare, P. V. Khadikar, Oxid. Commun.
33, (2010)
288

J. Chil. Chem. Soc., 55, Nº 3 (2010)
Model
Size
Coded
Variables
7
0.7389
0.7763
0.8240
0.8701
ABDGJPS
R2
8
ABCDGJPS
ABCDEHJPS
ABCDEFHJPS
1
2
0.3687
0.5385
0.6190
0.6797
0.7140
0.7646
0.8110
0.8255
0.8493
0.8580
D
9
DI
10
3
DIP
ADIS
Model
Size
Coded
Variables
R2
4
5
ADIPS
1
2
3
4
5
6
0.1118
0.3602
0.4866
0.5640
0.5719
0.5733
F
6
DIOPRS
EF
7
DIMOPRS
ADIMOPRS
ADILMNOPS
ADFILMNOPS
AEF
8
ADEF
ABDEF
ABCDEF
9
10
Model
Size
Coded
Variables
(3) E.coli
R2
(i)Using variables: C2, C9-C13. (ii) Using variables: C2, C14-C31.
(iii)Using variables: C2, C9-C31.
1
2
3
4
5
6
7
8
9
10
0.2691
0.3229
0.4946
0.5242
0.5737
0.6869
0.7707
0.7898
0.8118
0.8335
L
LS
Model
Size
Coded
Variables
R2
IPU
1
2
4
3
5
6
0.0790
0.3151
0.3378
0.3328
0.3400
0.3400
C
IPUV
EF
DIOPU
ABEF
AEF
AIJMOU
AIJMOUX
AIJMOQUX
ACIJMOQUX
ACIJMOQRUX
ABCEF
ABCDEF
Model
Size
Coded
Variables
R2
(5) K.pneumoniae
1
2
0.2784
U
(i)Using variables: C2, C9-C13 (ii) Using variables: C2, C14-C31 (iii)
Using variables: C2, C9-C31.
0.4101
0.5224
0.5922
0.6099
0.6294
0.7714
0.7836
0.7993
0.8047
AM
3
AMW
Model
Size
Coded
Variables
4
AMUW
R2
5
AJMUV
1
2
3
4
5
6
0.0058
0.0184
0.0474
0.0567
0.0597
0.0631
F
6
ABJMUV
BF
7
ABDEKUV
ABDEHKUV
ABDEGKRUW
ABCDEHKRUW
BCF
8
BCEF
ABCEF
ABCDEF
9
10
Model
Size
Coded
Variables
R2
Model
Size
Coded
Variables
R2
1
2
0.2784
0.4101
0.5224
0.5922
0.6099
0.6197
0.7214
0.7577
0.7877
0.8675
P
1
2
0.0807
0.1521
0.3034
0.4448
0.4753
0.4851
0.5093
0.6532
0.7020
0.7271
P
AH
DP
3
AHR
3
DEP
4
AHPR
4
GHJL
5
AEHPQ
5
GHJLP
6
AEHPRS
ADEGHJQ
ACDEGHJQ
ACDEGHJOQ
BCDEGHJLOQ
6
GHJLOP
FGHJLNP
BCEGHJOQ
BCEGHJOQR
BCEGHJNOQR
7
7
8
8
9
9
10
10
Model
Size
Coded
Variables
(4)S.aureus
R2
(i)Using variables: C2, C9-C13. (ii) Using variables: C2, C14-C31.
(iii)Using variables: C2, C9-C31.
1
2
3
4
5
6
7
8
9
10
0.0807
0.1603
0.3181
0.3884
0.4576
0.5170
0.5723
0.6142
0.7483
0.7604
U
CU
Model
Size
Coded
Variables
R2
CJU
1
2
3
4
5
6
0.2691
0.3229
0.4946
0.5242
0.6138
0.6514
G
CJMU
GN
CJMQU
DKP
CDEJMU
ACDEJMU
ACDEJMQU
CDGHLMTVX
CFGHLMTVWX
DKPQ
ADFGP
ADFGPQ
289

J. Chil. Chem. Soc., 55, Nº 3 (2010)
(6) P.aeruginosa
(i)Using variables: C2, C9-C13. (ii) Using variables: C2, C14-C31 (iii)Using variables: C2, C9-C31.
Model
Size
Coded
Variables
R2
1
2
3
4
5
6
7
8
9
10
0.1143
0.1830
0.2670
0.3463
0.3899
0.4040
0.4633
0.4846
0.5220
0.5293
B
BV
BJV
FMNV
EMNUW
EJNRUV
CIKNRUW
IKNRSTUX
CIKNRSTUX
CFIKNRSTU
Model
Size
Coded
Variables
R2
1
2
3
4
5
6
0.1758
0.2032
0.2085
0.2144
0.2181
0.2192
E
AF
ACD
ACDE
ACDEF
ABCDEF
Model
Size
Coded
Variables
R2
1
2
0.1031
0.1227
0.1395
0.3097
0.3603
0.4273
0.4579
0.4846
0.4978
0.5124
Q
QS
3
AQS
4
DFIM
5
DFIMQ
DFIMPR
DFIMNPR
DFIMNOPS
CDFIMNOPS
CDHIJMNOPS
6
7
8
9
10
(2). Correlation matrix for the parameters of most appropriate model for modeling antibacterial activity.
B.subtilis
C2
C3
C17
C20
C24
C27
C28
C30
C32
C2
1.0000
0.4219
0.4146
0.4270
0.3689
-0.0481
-0.1762
-0.1112
0.2509
0.4219
1.0000
0.9726
0.9745
0.9061
0.5369
0.2669
0.4923
0.1334
0.4146
0.9726
1.0000
0.9988
0.9153
0.4974
0.2964
0.4745
0.0777
0.4270
0.9745
0.9988
1.0000
0.9111
0.5021
0.2772
0.4702
0.0918
0.3689
0.9061
0.9153
0.9111
1.00001
0.4814
0.2737
0.4410
0.1228
-0.0481
0.5369
0.4974
0.5021
0.4814
1.0000
0.2610
0.8064
0.1670
-0.1762
0.2669
0.2964
0.2772
0.2737
0.2610
1.0000
0.1793
-0.0432
-0.1112
0.4923
0.4745
0.4702
0.44`0
0.8064
0.1793
1.0000
0.1906
0.2509
0.1334
0.0777
0.0918
0.1228
0.1670
-0.0432
0.1906
1.0000
C3
C17
C20
C24
C27
C28
C30
C32
S.typhi
C2
C16
C21
C24
C25
C26
C27
C28
C33
C2
1.0000
0.3080
0.3315
0.3688
0.3297
0.2483
-0.0481
-0.1762
-0.1436
0.3080
1.0000
0.9937
0.8849
0.9807
0.6253
0.5789
0.3521
0.6072
0.3315
0.9937
1.0000
0.9011
0.9943
0.6212
0.5378
0.3518
0.5575
0.3688
0.8849
0.9011
1.0000
0.9023
0.4646
0.4813
0.2736
0.5252
0.3297
0.9807
0.9943
0.9023
1.0000
0.6412
0.5009
0.3536
0.5323
0.2483
0.6253
0.6212
0.4646
0.6412
1.0000
0.1160
0.1982
0.2289
-0.0481
0.5789
0.5378
0.4813
0.5009
0.1160
1.0000
0.2609
0.3838
-0.1762
0.3521
0.3518
0.2736
0.3536
0.1982
0.2609
1.0000
0.4722
-0.1436
0.6072
0.5575
0.5252
0.5323
0.2289
0.3838
0.4722
1.0000
C16
C21
C24
C25
C26
C27
C28
C33
290

J. Chil. Chem. Soc., 55, Nº 3 (2010)
E.coli
C2
C9
C11
C12
C15
C18
C28
C29
C34
C2
C9
1.0000
0.6249
0.4232
0.4233
0.3286
0.4322
-0.1761
-0.1071
-0.4132
0.6249
1.0000
0.7893
0.7895
0.6337
0.7448
-0.0006
0.1882
-0.0267
0.4232
0.7893
1.0000
1.0000
0.9628
0.9900
0.2663
0.4913
0.2523
0.4233
0.7895
1.0000
1.0000
0.9627
0.9899
0.2662
0.4911
0.2523
0.3286
0.6337
0.9628
0.9627
1.0000
0.9791
0.3389
0.5314
0.2559
0.4322
0.7448
0.9900
0.9899
0.9791
1.0000
0.2765
0.4800
0.2080
-0.1762
-0.0007
0.2663
0.2662
0.3389
0.2765
1.0000
0.1830
0.9791
-0.1071
0.1882
0.4913
0.4911
0.5314
0.4800
0.1830
1.0000
1.0000
-0.4132
-0.0267
0.2523
0.2523
0.2559
0.2080
0.5276
0.0675
0.9791
C11
C12
C15
C18
C28
C29
C34
S.aureus
C2
C16
C17
C20
C22
C24
C28
C31
C35
C2
1.0000
0.3080
0.4145
0.4269
0.4237
0.3688
-0.1761
-0.0529
0.1430
0.3080
1.0000
0.9520
0.9467
0.9796
0.8849
0.3521
0.3764
0.3216
0.4145
0.9520
1.0000
0.9987
0.9729
0.9153
0.2963
0.2350
0.2868
0.4269
0.9467
0.9987
1.0000
0.9745
0.9110
0.2772
0.2215
0.2912
0.4237
0.9796
0.9729
0.9745
1.0000
0.8961
0.2782
0.2603
0.2652
0.3688
0.8849
0.9153
0.9110
0.8961
1.0000
0.2736
0.1839
0.2162
-0.1762
0.3521
0.2963
0.2772
0.2782
0.2736
1.0000
0.0927
-0.1530
-0.0529
0.3764
0.2350
0.2215
0.2603
0.1839
0.0927
1.0000
0.3182
0.1437
0.3216
0.2868
0.2912
0.2652
0.2162
-0.1530
0.3186
1.0000
C16
C17
C20
C22
C24
C28
C31
C35
K.pneumoniae
C2
C10
C11
C12
C17
C20
C24
C28
C36
C2
1.0000
0.2768
0.4232
0.4233
0.4145
0.4269
0.3688
-0.1761
0.1124
0.2764
1.0000
0.9609
0.9608
0.9426
0.9346
0.8934
0.3878
0.1498
0.4232
0.9609
1.0000
1.0000
0.9727
0.9746
0.9063
0.2663
0.1297
0.4233
0.9608
1.0000
1.0000
0.9727
0.9746
0.9063
0.2662
0.1297
0.4145
0.9426
0.9727
0.9727
1.0000
0.9987
0.9153
0.2963
0.0399
0.4269
0.9346
0.9746
0.9746
0.9987
1.0000
0.9110
0.2772
0.0504
0.3688
0.8934
0.9063
0.9063
0.9153
0.9110
1.0000
0.2736
0.1732
-0.1762
0.3878
0.2663
0.2662
0.2963
0.2772
0.2736
1.0000
-0.2846
0.1127
0.1498
0.1297
0.1297
0.0399
0.0504
0.1732
-0.2846
1.0000
C10
C11
C12
C17
C20
C24
C28
C36
P.aeruginosa
C16
C18
C21
C25
C26
C27
C28
C31
C37
C16
C18
C21
C25
C26
C27
C28
C31
C37
1.0000
0.9818
0.9937
0.9807
0.6253
0.5789
0.3521
0.3764
0.2268
0.9818
1.0000
0.9875
0.9768
0.5828
0.5147
0.2765
0.2871
0.2197
0.9937
0.9875
1.0000
0.9943
0.6212
0.5378
0.3518
0.3461
0.1836
0.9807
0.9768
0.9943
1.0000
0.6412
0.5009
0.3536
0.3479
0.1675
0.6253
0.5828
0.6212
0.6412
1.0000
0.1160
0.1982
0.7700
0.0159
0.5789
0.5147
0.5378
0.5009
0.1160
1.0000
0.2609
0.3795
0.3184
0.3521
0.2765
0.3518
0.3536
0.1982
0.2609
1.0000
0.0927
0.1042
0.3764
0.2871
0.3461
0.3479
0.7700
0.3795
0.0927
1.0000
0.1169
0.2268
0.2197
0.1836
0.1675
0.0159
0.3184
0.1042
0.1169
1.0000
291

J. Chil. Chem. Soc., 55, Nº 3 (2010)
(3) Ridge regression parameters: VIF, Tolerance, Eigen values,
Condition number for the most appropriate model for modeling antibacterial
activity.
K.pneumoniae
P.aeruginosa
Condition
Number
Variable
VIF
Tolerance Eigenvalue
B.subtilis S.typhi
C16
C18
C21
C25
C26
C27
C28
31.4770
29.4035
19.4545
24.2164
9.4800
0.0048
0.0127
0.0012
0.0046
0.0727
0.1756
0.5851
5.073812
1.206077
0.859504
0.788952
0.043209
0.018254
0.009341
1.00
Condition
Number
Variable VIF
Tolerance
Eigenvalue
4.21
5.90
C2
1.3667
0.6783
0.0048
0.0016
0.1558
0.0050
0.3649
0.4307
4.806702
1.288230
0.831546
0.588309
0.361899
0.108315
0.013966
1.00
6.43
C16
C21
C24
C25
C26
C27
24.7792
19.6221
5.7580
25.3768
2.0528
1.9469
3.73
117.42
277.95
543.20
5.78
4.1585
8.17
1.3646
13.28
44.38
344.16
Condition
Number
Variable
VIF
Tolerance
Eigenvalue
C2
1.5844
16.2188
7.0243
7.0662
13.4693
12.6196
6.0028
1.3852
0.5515
0.0165
0.0000
0.0000
0.0010
0.0010
0.1428
0.6056
6.019250
1.196181
0.552880
0.134921
0.069433
0.026828
0.000506
0.000000
1.00
Condition
Number
Variable
VIF
Tolerance
Eigenvalue
C10
C11
C12
C17
C20
C24
C28
5.03
10.89
C2
1.6190
0.5921
0.0446
0.0015
0.0015
0.1464
0.2633
0.0701
0.0366
4.758598
1.540658
0.930304
0.432212
0.185510
0.119560
0.002257
0.000001
1.00
44.61
C3
16.5952
16.0823
15.8720
5.7972
3.09
86.69
C17
C20
C24
C27
C24
C25
5.12
224.36
11896.45
94674753.04
11.01
25.65
3.0755
39.80
10.2254
17.6452
3148.50
6525191.27
E.coli S.aureus
Condition
Number
Variable
VIF
Tolerance
Eigenvalue
C2
1.5640
0.4156
0.0063
0.0009
0.0008
0.0049
0.1495
0.6546
5.131911
1.242010
0.923066
0.510967
0.131963
0.053840
0.005835
1.00
C16
C17
C20
C22
C24
C28
25.2418
14.0883
11.8939
28.7419
5.8384
4.13
5.56
10.04
38.89
95.32
879.56
1.3123
Condition
Number
1.00
Variable
VIF
Tolerance
Eigenvalue
C2
1.9409
5.7861
11.1715
11.2217
25.4340
36.8436
1.2886
1.6359
0.4508
0.1059
0.0000
0.0000
0.0181
0.0071
0.7392
0.5153
5.092069
1.466185
0.802052
0.366760
0.251692
0.015330
0.005911
0.000000
C9
3.47
C11
C12
C15
C18
C28
C29
6.35
13.88
20.23
332.15
861.46
91622650.97
292