A new version of ResearchHub is available.Try it now
Paper
Preprint
3
PIC50: An open source tool for interconversion of PIC50values and IC50for efficient data representation and analysis
Authors
Published
Oct 18, 2022
Show more
Save
Tip3
TipSave
PIC50: An open source tool for interconversion of PIC50 values and IC50 for efficient data
representation and analysis
Aman Thakur1, Ajay Kumar2, Vivek Sharma3, Vineet Mehta3*
1 DCO, Govt. of Rajasthan, Bharatpur, Rajasthan 321001
2 Institute of Pharmaceutical Sciences, Kurukshetra University, Haryana 136119
3 Department of Pharmacology, Govt. College of Pharmacy, Rohru, District Shimla, Himachal
Pradesh 171207
*Corresponding Author
Dr. Vineet Mehta
Assistant Professor (Pharmacology)
Department of Pharmacology
Govt. College of Pharmacy, Rohru
District Shimla, Himachal Pradesh 171207
E. mail: vineet.mehta20@gmail.com
Phone: 01781-241306
ORCID ID: https://orcid.org/0000-0003-0485-5076
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
representation and analysis
Aman Thakur1, Ajay Kumar2, Vivek Sharma3, Vineet Mehta3*
1 DCO, Govt. of Rajasthan, Bharatpur, Rajasthan 321001
2 Institute of Pharmaceutical Sciences, Kurukshetra University, Haryana 136119
3 Department of Pharmacology, Govt. College of Pharmacy, Rohru, District Shimla, Himachal
Pradesh 171207
*Corresponding Author
Dr. Vineet Mehta
Assistant Professor (Pharmacology)
Department of Pharmacology
Govt. College of Pharmacy, Rohru
District Shimla, Himachal Pradesh 171207
E. mail: vineet.mehta20@gmail.com
Phone: 01781-241306
ORCID ID: https://orcid.org/0000-0003-0485-5076
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
Abstract
Half-maximal inhibitory concentration (IC50) is used to determine the potency of a drug against a
variety of enzymes/ biological targets associated with the pathogenesis of multiple disorders. The
IC50 values can be depicted in multiple ways, which makes it difficult to analyze the results
presented in different concentrations. Representing data in the form of PIC50 values depicting the
IC50 values as the negative logarithm of IC50 in molar concentration is considered to be a better
approach as it not only makes data easily understandable but also eliminates the possibility of
errors in data representation and reproducibility. Considering the importance of data
representation for a better understanding of data and comparing efficacy and potency of the
drugs, besides, the significance of PIC50 value in the field of CADD, we found that at present
there is no single open-source software available to convert the IC50 values to PIC50 values and
vice versa from millimolar to picomolar range. Therefore, in the present study, we develop a tool
that could help researchers to interconvert IC50 values and PIC50 values in a reliable way to
eliminate the possibility of errors. We validated our tool through three case studies where the
data generated by our tool was found to be 100% accurate. Moreover, we present a case where
data was published in literature with errors in calculated PIC50 values and demonstrated the
importance and reliability of our tool.
Keywords: PIC50 value calculator; PIC50; IC50 value calculator; IC50; QSAR; CADD.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
Half-maximal inhibitory concentration (IC50) is used to determine the potency of a drug against a
variety of enzymes/ biological targets associated with the pathogenesis of multiple disorders. The
IC50 values can be depicted in multiple ways, which makes it difficult to analyze the results
presented in different concentrations. Representing data in the form of PIC50 values depicting the
IC50 values as the negative logarithm of IC50 in molar concentration is considered to be a better
approach as it not only makes data easily understandable but also eliminates the possibility of
errors in data representation and reproducibility. Considering the importance of data
representation for a better understanding of data and comparing efficacy and potency of the
drugs, besides, the significance of PIC50 value in the field of CADD, we found that at present
there is no single open-source software available to convert the IC50 values to PIC50 values and
vice versa from millimolar to picomolar range. Therefore, in the present study, we develop a tool
that could help researchers to interconvert IC50 values and PIC50 values in a reliable way to
eliminate the possibility of errors. We validated our tool through three case studies where the
data generated by our tool was found to be 100% accurate. Moreover, we present a case where
data was published in literature with errors in calculated PIC50 values and demonstrated the
importance and reliability of our tool.
Keywords: PIC50 value calculator; PIC50; IC50 value calculator; IC50; QSAR; CADD.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
Introduction
Half-maximal inhibitory concentration (IC50) is one of the extensively acknowledged parameters
to assess the potency of a drug against a variety of enzymes/ biological targets that are associated
with the pathogenesis of disorders like Alzheimer’s disease, depression, diabetes mellitus,
cancer, etc. (Aykul and Martinez-Hackert, et al., 2016). IC50 is a measure of the potency of a
drug and is defined as the minimum concentration obligatory for inhibiting the biological activity
of a target by 50% (Aykul and Martinez-Hackert, et al., 2016). Literature reports suggest that the
IC50 values of a particular drug against a particular enzymes/ biological targets can be depicted in
multiple ways, which makes it difficult to analyze the results and reach a decisive conclusion
regarding the exact efficacy and potency of a drug (Changyong et al., 2014). One of the better
ways to present data of a scientific finding is in a logarithmic form as it spreads out the data in
such a way that the shape of the curve and the quality of the fit are readily visible when the
concentrations cover a wide range. There is a group of scientists that argue that linear functions
can be used for percent inhibition but should not be used to calculate IC50 as the IC50 curve is not
a linear function, rather it is a saturating function. Representing data in the form of PIC50 values
is considered to be a better approach as it not only makes data easily understandable but also
eliminates the possibility of errors in data representation and reproducibility (Changyong et al.,
2014). PIC50 is the approach for depicting the IC50 values as the negative logarithm of IC50 in
molar concentration, therefore it makes data more convenient for the readers to understand and
compare the potency of different drugs at the same molar levels (Abdulrahman et al., 2021;
Thakur et al., 2022). PIC50 values are now being extensively used in an array of computer-aided
drug designing (CADD) approaches such as Quantitative Structural Activity Relationship
(QSAR) (Hadaji et al., 2017; Ramalakshmi et al., 2021), comparative molecular field analysis
(coMFA), comparative molecular similarity indices analysis (coMSIA) (Liang et al., 2013),
Pharmacophore modeling (Vyas et al., 2013), etc. and the outcome of these tools solely depends
on the reliability of input data of PIC50 values. As the IC50 value is distributed on a wider range,
hence to alter these values to a normal distribution, the PIC50 value is used so that a vast
numerical distribution can be represented in certain intervals (da Silva Costa et al., 2018; Ferreira
et al., 2019). Moreover, it also endows with a logarithmic approach to the study as higher values
of PIC50 signify higher potency (Hendrickx et al., 2018). Although the approach to convert IC50
values to PIC50 values seems to be simple, through rigorous literature review we found that it is a
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
Half-maximal inhibitory concentration (IC50) is one of the extensively acknowledged parameters
to assess the potency of a drug against a variety of enzymes/ biological targets that are associated
with the pathogenesis of disorders like Alzheimer’s disease, depression, diabetes mellitus,
cancer, etc. (Aykul and Martinez-Hackert, et al., 2016). IC50 is a measure of the potency of a
drug and is defined as the minimum concentration obligatory for inhibiting the biological activity
of a target by 50% (Aykul and Martinez-Hackert, et al., 2016). Literature reports suggest that the
IC50 values of a particular drug against a particular enzymes/ biological targets can be depicted in
multiple ways, which makes it difficult to analyze the results and reach a decisive conclusion
regarding the exact efficacy and potency of a drug (Changyong et al., 2014). One of the better
ways to present data of a scientific finding is in a logarithmic form as it spreads out the data in
such a way that the shape of the curve and the quality of the fit are readily visible when the
concentrations cover a wide range. There is a group of scientists that argue that linear functions
can be used for percent inhibition but should not be used to calculate IC50 as the IC50 curve is not
a linear function, rather it is a saturating function. Representing data in the form of PIC50 values
is considered to be a better approach as it not only makes data easily understandable but also
eliminates the possibility of errors in data representation and reproducibility (Changyong et al.,
2014). PIC50 is the approach for depicting the IC50 values as the negative logarithm of IC50 in
molar concentration, therefore it makes data more convenient for the readers to understand and
compare the potency of different drugs at the same molar levels (Abdulrahman et al., 2021;
Thakur et al., 2022). PIC50 values are now being extensively used in an array of computer-aided
drug designing (CADD) approaches such as Quantitative Structural Activity Relationship
(QSAR) (Hadaji et al., 2017; Ramalakshmi et al., 2021), comparative molecular field analysis
(coMFA), comparative molecular similarity indices analysis (coMSIA) (Liang et al., 2013),
Pharmacophore modeling (Vyas et al., 2013), etc. and the outcome of these tools solely depends
on the reliability of input data of PIC50 values. As the IC50 value is distributed on a wider range,
hence to alter these values to a normal distribution, the PIC50 value is used so that a vast
numerical distribution can be represented in certain intervals (da Silva Costa et al., 2018; Ferreira
et al., 2019). Moreover, it also endows with a logarithmic approach to the study as higher values
of PIC50 signify higher potency (Hendrickx et al., 2018). Although the approach to convert IC50
values to PIC50 values seems to be simple, through rigorous literature review we found that it is a
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
tricky process and there are significant errors in the calculated PIC50 values in published research
papers.
Considering the importance of data representation for a better understanding of data and
comparing efficacy and potency of the drugs, besides, the significance of PIC50 value in the field
of CADD, we found that at present there is no single open-source software available to convert
the IC50 values to PIC50 values and vice versa from milimolar to picomolar range. Therefore, in
the present study, we aimed to develop a tool that can help researchers to interconvert IC50
values and PIC50 values in a reliable way to eliminate the possibility of errors. This software will
not only reduce the time for calculating the PIC50 or IC50 values but significantly eliminates the
chance of human error in calculation.
Material and Methods
Developed PIC50 Tool
The current PIC50 tool can be downloaded free from
https://www.researchgate.net/publication/363769944_PIC50_to_IC50_and_IC50_to_PIC50_calc
ulator#fullTextFileContent. This tool allows user to calculate PIC50 values form the IC50 values
depicted in millimolar, micromolar, nanomolar, or picomolar concentration range.
Software and its functionalities
The current software “PIC50” is developed on python programming language and further
converted to .exe file format for ease of portability. It works at ease on and above Windows 7
platforms. Currently, it consists of two modules: (i) Calculating IC50 from PIC50 value and (ii)
Calculating PIC50 from IC50 value.
After opening the .exe file, the user will get two options for calculating IC50 or PIC50, enter “1”
for calculating the IC50 value from PIC50 and “2” for calculating the PIC50 value from IC50 value.
Afterward, the user will get further options to calculate their values either in millimolar,
micromolar, nanomolar, or picomolar range as per the requirement. Users will be prompted to
enter “3” for millimolar, “6” for micromolar, “9” for nanomolar, and “12” for picomolar values.
After entering the IC50/PIC50 values user will simply get their respective PIC50/IC50 values
(Image 1). There is no limit for calculating the number of values as a user can calculate values
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
papers.
Considering the importance of data representation for a better understanding of data and
comparing efficacy and potency of the drugs, besides, the significance of PIC50 value in the field
of CADD, we found that at present there is no single open-source software available to convert
the IC50 values to PIC50 values and vice versa from milimolar to picomolar range. Therefore, in
the present study, we aimed to develop a tool that can help researchers to interconvert IC50
values and PIC50 values in a reliable way to eliminate the possibility of errors. This software will
not only reduce the time for calculating the PIC50 or IC50 values but significantly eliminates the
chance of human error in calculation.
Material and Methods
Developed PIC50 Tool
The current PIC50 tool can be downloaded free from
https://www.researchgate.net/publication/363769944_PIC50_to_IC50_and_IC50_to_PIC50_calc
ulator#fullTextFileContent. This tool allows user to calculate PIC50 values form the IC50 values
depicted in millimolar, micromolar, nanomolar, or picomolar concentration range.
Software and its functionalities
The current software “PIC50” is developed on python programming language and further
converted to .exe file format for ease of portability. It works at ease on and above Windows 7
platforms. Currently, it consists of two modules: (i) Calculating IC50 from PIC50 value and (ii)
Calculating PIC50 from IC50 value.
After opening the .exe file, the user will get two options for calculating IC50 or PIC50, enter “1”
for calculating the IC50 value from PIC50 and “2” for calculating the PIC50 value from IC50 value.
Afterward, the user will get further options to calculate their values either in millimolar,
micromolar, nanomolar, or picomolar range as per the requirement. Users will be prompted to
enter “3” for millimolar, “6” for micromolar, “9” for nanomolar, and “12” for picomolar values.
After entering the IC50/PIC50 values user will simply get their respective PIC50/IC50 values
(Image 1). There is no limit for calculating the number of values as a user can calculate values
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
‘n’ times from the software. The current software uses the following equations for calculating the
IC50 and PIC50 values respectively:
IC50 = 10(x - PIC50), and PIC50 = x - log10 (IC50)
Where, x = 3 for millimolar, 6 for micromolar, 9 for nanomolar, and 12 for picomolar
concentrations
Image 1: Picture showing the calculation of PIC50 in the micromolar range
Results and Discussion
To confirm the functionality of the current software we performed three case studies. In the first
case study, a series of 30 1H Pyrazole-1-carbothioamide derivatives were employed to develop
a QSAR model, and based on this model PIC50 value (in micromolar) of novel 11 derivatives
was calculated. In our study, we predicted accurately the PIC50 of 30 derivatives used and also
converted the PIC50 to IC50 of 11 compounds, whose activity was predicted for comparative
study with the 30 derivatives (Hajalsiddig et al., 2020). Results are depicted in Table 1 (a, b). In
this study, author calculated experimental PIC50 value of compound “C16” as 5.24 whereas
correct experimental PIC50 value of the same is calculated from our tool is 5.27. This error in
calculating PIC50 value of this compound also resulted in wrong interpretation of residuals
(Experimental PIC50; Predicted PIC50) from the two QSAR equations generated in the study. In
the equation 1, author has calculated residual value of compound “C16” as 0.12, however, it
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
IC50 and PIC50 values respectively:
IC50 = 10(x - PIC50), and PIC50 = x - log10 (IC50)
Where, x = 3 for millimolar, 6 for micromolar, 9 for nanomolar, and 12 for picomolar
concentrations
Image 1: Picture showing the calculation of PIC50 in the micromolar range
Results and Discussion
To confirm the functionality of the current software we performed three case studies. In the first
case study, a series of 30 1H Pyrazole-1-carbothioamide derivatives were employed to develop
a QSAR model, and based on this model PIC50 value (in micromolar) of novel 11 derivatives
was calculated. In our study, we predicted accurately the PIC50 of 30 derivatives used and also
converted the PIC50 to IC50 of 11 compounds, whose activity was predicted for comparative
study with the 30 derivatives (Hajalsiddig et al., 2020). Results are depicted in Table 1 (a, b). In
this study, author calculated experimental PIC50 value of compound “C16” as 5.24 whereas
correct experimental PIC50 value of the same is calculated from our tool is 5.27. This error in
calculating PIC50 value of this compound also resulted in wrong interpretation of residuals
(Experimental PIC50; Predicted PIC50) from the two QSAR equations generated in the study. In
the equation 1, author has calculated residual value of compound “C16” as 0.12, however, it
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
should be 0.15. Likewise, in equation 2, the value of the same was reported to be 0.9 but it
should be 0.12. This error will be automatically be carried forward in the whole QSAR model
generation and will affect the overall accuracy of the QSAR model significantly.
Table 1a: Reported PIC50 values of the 30 compounds (Hajalsiddig et al., 2020) and PIC50 values
predicted by our tool
S. No. Name of Compound PIC50 calculated by the authors PIC50 calculated by our tool
1 C1 6.08 6.081
2 C2 5.87 5.866
3 C3 5.66 5.665
4 C4 6.47 6.468
5 C5 7.15 7.155
6 C6 6.89 6.886
7 C7 5.51 5.514
8 C8 5.28 5.285
9 C9 5.41 5.412
10 C10 5.37 5.375
11 C11 5.20 5.203
12 C12 5.14 5.135
13 C13 5.16 5.164
14 C14 5.13 5.132
15 C15 5.24 5.241
16 C16 5.24 5.278
17 C17 5.11 5.109
18 C18 5.18 5.183
19 C19 5.13 5.137
20 C20 5.19 5.191
21 C21 5.04 5.050
22 C22 5.08 5.089
23 C23 4.99 4.997
24 C24 5.00 5.007
25 C25 5.09 5.092
26 C26 4.95 4.948
27 C27 4.97 4.973
28 C28 4.87 4.874
29 C29 4.91 4.911
30 C30 4.95 4.960
Table 1b: Reported PIC50 values of the 11 compounds (Hajalsiddig et al., 2020) and PIC50 values
predicted by our tool
S. No. Name of Compound PIC50 calculated by the authors PIC50 calculated by our tool
1 C1 6.08 6.081
2 C2 5.87 5.866
3 C3 5.66 5.665
4 C4 6.47 6.468
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
should be 0.12. This error will be automatically be carried forward in the whole QSAR model
generation and will affect the overall accuracy of the QSAR model significantly.
Table 1a: Reported PIC50 values of the 30 compounds (Hajalsiddig et al., 2020) and PIC50 values
predicted by our tool
S. No. Name of Compound PIC50 calculated by the authors PIC50 calculated by our tool
1 C1 6.08 6.081
2 C2 5.87 5.866
3 C3 5.66 5.665
4 C4 6.47 6.468
5 C5 7.15 7.155
6 C6 6.89 6.886
7 C7 5.51 5.514
8 C8 5.28 5.285
9 C9 5.41 5.412
10 C10 5.37 5.375
11 C11 5.20 5.203
12 C12 5.14 5.135
13 C13 5.16 5.164
14 C14 5.13 5.132
15 C15 5.24 5.241
16 C16 5.24 5.278
17 C17 5.11 5.109
18 C18 5.18 5.183
19 C19 5.13 5.137
20 C20 5.19 5.191
21 C21 5.04 5.050
22 C22 5.08 5.089
23 C23 4.99 4.997
24 C24 5.00 5.007
25 C25 5.09 5.092
26 C26 4.95 4.948
27 C27 4.97 4.973
28 C28 4.87 4.874
29 C29 4.91 4.911
30 C30 4.95 4.960
Table 1b: Reported PIC50 values of the 11 compounds (Hajalsiddig et al., 2020) and PIC50 values
predicted by our tool
S. No. Name of Compound PIC50 calculated by the authors PIC50 calculated by our tool
1 C1 6.08 6.081
2 C2 5.87 5.866
3 C3 5.66 5.665
4 C4 6.47 6.468
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
5 C5 7.15 7.155
6 C6 6.89 6.886
7 C7 5.51 5.514
8 C8 5.28 5.285
9 C9 5.41 5.412
10 C10 5.37 5.375
11 C11 5.20 5.203
In the second study, 28 berberine analogs were used to develop a QSAR model against coxsackie
B1 virus and PIC50 values of all the analogs were measured in micromolar. We also used our
software and predicted the PIC50 values accurately and in a very short time period (Obadawo et
al., 2022). Results are depicted in Table 2. In this study, author has calculated the PIC50 value of
only 24 compounds out of 28 compounds mentioned in the literature. The PIC50 values of
Compound No. “4”, “19”, “26” and “27” was not calculated by the author. Our tool successfully
calculated the PIC50 of these excluded compounds accurately. This suggest that the tool
developed in the present study is efficient to calculate PIC50 values over wide range of
concentrations reliably and accurately.
Table 2: Reported PIC50 values of the 28 compounds (Obadawo et al., 2022) and PIC50 values
predicted by our tool
S. No. Name of Compound PIC50 calculated by author PIC50 calculated by our tool
1 1 5.644 5.644
2 2 5.4225 5.422
3 3 5.2628 5.262
4 4 Not Calculated by author 5.634
5 5 4.9066 4.907
6 6 4.8665 4.866
7 7 5.1918 5.192
8 8 5.2984 5.298
9 9 5.5058 5.506
10 10 5.5918 5.592
11 11 4.8665 4.866
12 12 5.2097 5.210
13 13 4.983 4.983
14 14 5.2182 5.218
15 15 5.3019 5.302
16 16 5.0353 5.035
17 17 4.6799 4.680
18 18 4.9626 4.963
19 19 Not calculated by the author 4.487
20 20 4.3925 4.392
21 21 5.0353 5.035
22 22 4.9066 4.907
23 23 4.5406 4.541
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
6 C6 6.89 6.886
7 C7 5.51 5.514
8 C8 5.28 5.285
9 C9 5.41 5.412
10 C10 5.37 5.375
11 C11 5.20 5.203
In the second study, 28 berberine analogs were used to develop a QSAR model against coxsackie
B1 virus and PIC50 values of all the analogs were measured in micromolar. We also used our
software and predicted the PIC50 values accurately and in a very short time period (Obadawo et
al., 2022). Results are depicted in Table 2. In this study, author has calculated the PIC50 value of
only 24 compounds out of 28 compounds mentioned in the literature. The PIC50 values of
Compound No. “4”, “19”, “26” and “27” was not calculated by the author. Our tool successfully
calculated the PIC50 of these excluded compounds accurately. This suggest that the tool
developed in the present study is efficient to calculate PIC50 values over wide range of
concentrations reliably and accurately.
Table 2: Reported PIC50 values of the 28 compounds (Obadawo et al., 2022) and PIC50 values
predicted by our tool
S. No. Name of Compound PIC50 calculated by author PIC50 calculated by our tool
1 1 5.644 5.644
2 2 5.4225 5.422
3 3 5.2628 5.262
4 4 Not Calculated by author 5.634
5 5 4.9066 4.907
6 6 4.8665 4.866
7 7 5.1918 5.192
8 8 5.2984 5.298
9 9 5.5058 5.506
10 10 5.5918 5.592
11 11 4.8665 4.866
12 12 5.2097 5.210
13 13 4.983 4.983
14 14 5.2182 5.218
15 15 5.3019 5.302
16 16 5.0353 5.035
17 17 4.6799 4.680
18 18 4.9626 4.963
19 19 Not calculated by the author 4.487
20 20 4.3925 4.392
21 21 5.0353 5.035
22 22 4.9066 4.907
23 23 4.5406 4.541
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
24 24 4.8268 4.827
25 25 5.1681 5.168
26 26 Not calculated by the author 4.866
27 27 Not calculated by the author 4.801
28 28 5.1662 5.166
In the third case, 18 compounds were employed for developing a QSAR model for predicting
anti-malarial activity. The PIC50 values (in the nanomolar range) were used for performing the
QSAR study. The same was predicted by our software accurately (Sahu et al., 2014). Results are
depicted in Table 3. In this study two QSAR models were generated, one against Chloroquine
sensitive Plasmodioum falciparum strain (HB3) and second for Chloroquine resistant
Plasmodioum falciparum strain (Dd2). In the PIC50 value (in nanomolar range), author has
calculated the PIC50 value of compound “16b” as 7.119, however, the accurate value of the same
should be 7.286 that was predicted accurately by the tool developed in present study. Moreover,
for compound “17a” the PIC50 value calculated by author was 7.286 whereas the same should be
7.119, as calculated by our tool. This error will be automatically be carried forward in the whole
QSAR model generation and will affect the overall accuracy of the QSAR model significantly.
Table 3: Reported PIC50 values of the 18 compounds (Sahu et al., 2014) and PIC50 values predicted
by our tool
S. No. Name Of
Compound PIC50 calculated by author PIC50 calculated by our tool
1 4a 6.889 6.889
2 4b 7.249 7.249
3 4c 6.77 6.770
4 4d 6.987 6.987
5 4e 6.57 6.570
6 5a 7.506 7.506
7 5b 7.551 7.551
8 5c 7.073 7.073
9 5d 7.363 7.363
10 5e 6.562 6.562
11 6a 6.893 6.893
12 6b 7.001 7.001
13 7a 6.055 6.055
14 7b 5.593 5.593
15 16a 7.097 7.097
16 16b 7.119 7.286
17 17a 7.286 7.119
18 17b 7.121 7.121
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
25 25 5.1681 5.168
26 26 Not calculated by the author 4.866
27 27 Not calculated by the author 4.801
28 28 5.1662 5.166
In the third case, 18 compounds were employed for developing a QSAR model for predicting
anti-malarial activity. The PIC50 values (in the nanomolar range) were used for performing the
QSAR study. The same was predicted by our software accurately (Sahu et al., 2014). Results are
depicted in Table 3. In this study two QSAR models were generated, one against Chloroquine
sensitive Plasmodioum falciparum strain (HB3) and second for Chloroquine resistant
Plasmodioum falciparum strain (Dd2). In the PIC50 value (in nanomolar range), author has
calculated the PIC50 value of compound “16b” as 7.119, however, the accurate value of the same
should be 7.286 that was predicted accurately by the tool developed in present study. Moreover,
for compound “17a” the PIC50 value calculated by author was 7.286 whereas the same should be
7.119, as calculated by our tool. This error will be automatically be carried forward in the whole
QSAR model generation and will affect the overall accuracy of the QSAR model significantly.
Table 3: Reported PIC50 values of the 18 compounds (Sahu et al., 2014) and PIC50 values predicted
by our tool
S. No. Name Of
Compound PIC50 calculated by author PIC50 calculated by our tool
1 4a 6.889 6.889
2 4b 7.249 7.249
3 4c 6.77 6.770
4 4d 6.987 6.987
5 4e 6.57 6.570
6 5a 7.506 7.506
7 5b 7.551 7.551
8 5c 7.073 7.073
9 5d 7.363 7.363
10 5e 6.562 6.562
11 6a 6.893 6.893
12 6b 7.001 7.001
13 7a 6.055 6.055
14 7b 5.593 5.593
15 16a 7.097 7.097
16 16b 7.119 7.286
17 17a 7.286 7.119
18 17b 7.121 7.121
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
In another case study, a total of twenty-one 3-Iodochrome derivatives were synthesized and the
QSAR model was generated for potential fungicides (Kaushik et al., 2021). This study is a
perfect example of incorrect data representation arising due to a lack of a proper tool to calculate
PIC50 values from the IC50 values. In this study, ED50 values of the compounds were evaluated
against Sclerotium rolfsii in the mgL-1 (Kaushik et al., 2021). For developing the QSAR model,
the author utilized the pED50 values, however, for these calculations authors did not consider the
factor of the unit of measurement i.e. mg (10-3). We calculated the PIC50 values from our
software and compared them to the reported values. Table 4 depicts the pED50 value calculated
by the author (Kaushik et al., 2021) and the pED50 value calculated by our software from the
given ED50.
Table 4: Difference in the reported pED50 and pED50 values calculated by our software
S. No. Compound
Name ED50 (mgL-1)
pED50
(Calculated by the
author)
pED50
(Calculated by our
software)
1 4a 50.38 -1.70 1.30
2 4b 19.12 -1.28 1.72
3 4c 30.2 -1.48 1.52
4 4d 39.28 -1.59 1.41
5 4e 47.59 -1.68 1.32
6 4f 61.23 -1.79 1.21
7 4g 64.02 -1.81 1.19
8 4h 75.58 -1.88 1.12
9 4i 93.17 -1.97 1.03
10 4j 124.7 -2.10 0.90
11 4k 165.7 -2.22 0.78
12 4l 248.6 -2.40 0.60
13 4m 325 -2.51 0.49
14 4n 410 -2.61 0.39
15 4o 515.4 -2.71 0.29
16 4p 72.86 -1.86 1.14
17 4q 63.75 -1.80 1.20
18 4r 8.43 -0.93 2.07
19 4s 147.8 -2.17 0.83
20 4t 33.98 -1.53 1.47
21 4u 110.9 -2.04 0.96
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
QSAR model was generated for potential fungicides (Kaushik et al., 2021). This study is a
perfect example of incorrect data representation arising due to a lack of a proper tool to calculate
PIC50 values from the IC50 values. In this study, ED50 values of the compounds were evaluated
against Sclerotium rolfsii in the mgL-1 (Kaushik et al., 2021). For developing the QSAR model,
the author utilized the pED50 values, however, for these calculations authors did not consider the
factor of the unit of measurement i.e. mg (10-3). We calculated the PIC50 values from our
software and compared them to the reported values. Table 4 depicts the pED50 value calculated
by the author (Kaushik et al., 2021) and the pED50 value calculated by our software from the
given ED50.
Table 4: Difference in the reported pED50 and pED50 values calculated by our software
S. No. Compound
Name ED50 (mgL-1)
pED50
(Calculated by the
author)
pED50
(Calculated by our
software)
1 4a 50.38 -1.70 1.30
2 4b 19.12 -1.28 1.72
3 4c 30.2 -1.48 1.52
4 4d 39.28 -1.59 1.41
5 4e 47.59 -1.68 1.32
6 4f 61.23 -1.79 1.21
7 4g 64.02 -1.81 1.19
8 4h 75.58 -1.88 1.12
9 4i 93.17 -1.97 1.03
10 4j 124.7 -2.10 0.90
11 4k 165.7 -2.22 0.78
12 4l 248.6 -2.40 0.60
13 4m 325 -2.51 0.49
14 4n 410 -2.61 0.39
15 4o 515.4 -2.71 0.29
16 4p 72.86 -1.86 1.14
17 4q 63.75 -1.80 1.20
18 4r 8.43 -0.93 2.07
19 4s 147.8 -2.17 0.83
20 4t 33.98 -1.53 1.47
21 4u 110.9 -2.04 0.96
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
Considering the example of compound ‘4a’ from the table, ED50 was reported to be 50.38 mgL-1.
The authors computed the negative logarithm of ED50 values directly of the compound ‘4a’
without taking the 10-3 (for milligram) into the calculation. Since pED50 values are the negative
logarithm of EC50 values in molar concentration, the concentration factor has to be taken into
consideration. From our tool, pED50 of the same molecules was calculated by taking the unit of
milligram also into consideration and the values are depicted in Table 4. Interestingly, our results
demonstrated significant variations from the reported values. Our software utilized the following
calculation for calculating the pED50 values:
pED50 = - log(50.38 X 10-3) = 1.30 (For compound ‘4a’)
It is evident from Table 4 that our tool calculated the pED50 values accurately and also solved the
purpose of using the pED50 value, which is the logarithmic approach (higher values for more
potent compounds).
From our study, we efficiently calculated PIC50 values from IC50 values for a total of 108
compounds that were randomly selected from literature. We also identified the error in data
presentations in these studies, which was carried forward in the whole QSAR model generation
and had adversely affected the overall accuracy of the QSAR model significantly. Moreover, in
the last case study, entire calculations were found to have error in the determination of PIC50
values, which were predicted accurately by our tool.
Conclusion
QSAR is a well-established technique that has a wide application in the field of drug designing,
pharmaceuticals, ecotoxicity of industrial chemicals, materials science, etc. (Puzyn et al., 2010;
Roy et al., 2017). The accurate depiction of the PIC50 values is a crucial step for developing
robust and reliable QSAR models. Currently, there is no tool available that can accurately and
reliably calculate the PIC50 values and there are significant chances of errors in manual
methods. In our study, we have presented a user-friendly, simple, and convenient PIC50 value
calculator that can accurately calculate PIC50 values to eliminate any possibility of errors and
present error-free data in an easily understandable and reliable manner. This software is available
in open source
(https://www.researchgate.net/publication/363769944_PIC50_to_IC50_and_IC50_to_PIC50_cal
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
The authors computed the negative logarithm of ED50 values directly of the compound ‘4a’
without taking the 10-3 (for milligram) into the calculation. Since pED50 values are the negative
logarithm of EC50 values in molar concentration, the concentration factor has to be taken into
consideration. From our tool, pED50 of the same molecules was calculated by taking the unit of
milligram also into consideration and the values are depicted in Table 4. Interestingly, our results
demonstrated significant variations from the reported values. Our software utilized the following
calculation for calculating the pED50 values:
pED50 = - log(50.38 X 10-3) = 1.30 (For compound ‘4a’)
It is evident from Table 4 that our tool calculated the pED50 values accurately and also solved the
purpose of using the pED50 value, which is the logarithmic approach (higher values for more
potent compounds).
From our study, we efficiently calculated PIC50 values from IC50 values for a total of 108
compounds that were randomly selected from literature. We also identified the error in data
presentations in these studies, which was carried forward in the whole QSAR model generation
and had adversely affected the overall accuracy of the QSAR model significantly. Moreover, in
the last case study, entire calculations were found to have error in the determination of PIC50
values, which were predicted accurately by our tool.
Conclusion
QSAR is a well-established technique that has a wide application in the field of drug designing,
pharmaceuticals, ecotoxicity of industrial chemicals, materials science, etc. (Puzyn et al., 2010;
Roy et al., 2017). The accurate depiction of the PIC50 values is a crucial step for developing
robust and reliable QSAR models. Currently, there is no tool available that can accurately and
reliably calculate the PIC50 values and there are significant chances of errors in manual
methods. In our study, we have presented a user-friendly, simple, and convenient PIC50 value
calculator that can accurately calculate PIC50 values to eliminate any possibility of errors and
present error-free data in an easily understandable and reliable manner. This software is available
in open source
(https://www.researchgate.net/publication/363769944_PIC50_to_IC50_and_IC50_to_PIC50_cal
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
culator/stats) and can be easily downloaded and used. Our software not only calculates the PIC50
value from the IC50 value but also performs the IC50 calculations from the PIC50 value. The
developed PIC50 tool is capable of calculating values in all ranges of concentrations including
millimolar, micromolar, nanomolar, and picomolar, which makes this tool more valuable.
Further, to evaluate the functionality of our tool we examined three case studies and found all the
PIC50 and IC50 values were calculated with 100% accuracy and in very less time. Moreover, our
results (Table 1 to Table 4) demonstrate how there are vast possibilities of errors in data
representation and how that can be eliminated by using the developed PIC50 tool. For any
research, the accuracy of the data and time saving are the two most important criteria. Our
software provides both for all the researchers who used these parameters for their studies on an
open-source platform.
Conflict of Interest
The authors are not having any conflict of interest concerning any part of this study.
Acknowledgement
The authors acknowledge Govt. College of Pharmacy, Rohru and Institute of Pharmaceutical
Science, Kurukshetra University, Haryana for providing the facilities to conduct this study.
Author’s Contribution
AT contributed to designing the study, performed QSAR, and prepared the first draft of the
manuscript. AK and VS contributed to the study designing, technical inputs, and manuscript
editing. VM contributed to designing the study, Results analysis, manuscript preparation, and
editing analysis of the results.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
value from the IC50 value but also performs the IC50 calculations from the PIC50 value. The
developed PIC50 tool is capable of calculating values in all ranges of concentrations including
millimolar, micromolar, nanomolar, and picomolar, which makes this tool more valuable.
Further, to evaluate the functionality of our tool we examined three case studies and found all the
PIC50 and IC50 values were calculated with 100% accuracy and in very less time. Moreover, our
results (Table 1 to Table 4) demonstrate how there are vast possibilities of errors in data
representation and how that can be eliminated by using the developed PIC50 tool. For any
research, the accuracy of the data and time saving are the two most important criteria. Our
software provides both for all the researchers who used these parameters for their studies on an
open-source platform.
Conflict of Interest
The authors are not having any conflict of interest concerning any part of this study.
Acknowledgement
The authors acknowledge Govt. College of Pharmacy, Rohru and Institute of Pharmaceutical
Science, Kurukshetra University, Haryana for providing the facilities to conduct this study.
Author’s Contribution
AT contributed to designing the study, performed QSAR, and prepared the first draft of the
manuscript. AK and VS contributed to the study designing, technical inputs, and manuscript
editing. VM contributed to designing the study, Results analysis, manuscript preparation, and
editing analysis of the results.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
References
1. Abdulrahman HL, Uzairu A, Uba S. Computer modeling of some anti-breast cancer
compounds. Structural Chemistry. 2021; 32(2):679-87.
2. Aykul S, Martinez-Hackert E. Determination of half-maximal inhibitory concentration
using biosensor-based protein interaction analysis. Analytical biochemistry. 2016;
508:97-103.
3. Changyong FE, Hongyue WA, Naiji LU, Tian CH, Hua HE, Ying LU. Log-
transformation and its implications for data analysis. Shanghai archives of psychiatry.
2014; 26(2):105.
4. da Silva Costa J, da Silva Lopes Costa K, Cruz JV, da Silva Ramos R, Silva LB, Do
Socorro Barros Brasil D, de Paula da Silva CH, Dos Santos CB, da Cruz Macedo WJ.
Virtual screening and statistical analysis in the design of new caffeine analogues
molecules with potential epithelial anticancer activity. Current pharmaceutical design.
2018; 24(5):576-94.
5. Ferreira EF, Silva LB, Costa GV, Costa JS, Fujishima MA, Leão RP, Ferreira AL,
Federico LB, Silva CH, Rosa JM, Macêdo WJ. Identification of new inhibitors with
potential antitumor activity from polypeptide structures via hierarchical virtual screening.
Molecules. 2019; 24(16):2943.
6. Hadaji EG, Bourass M, Ouammou A, Bouachrine M. 3D-QSAR models to predict anti-
cancer activity on a series of protein P38 MAP kinase inhibitors. Journal of Taibah
University for Science. 2017; 11(3):392-407.
7. Hajalsiddig TT, Osman AB, Saeed AE. 2D-QSAR modeling and molecular docking
studies on 1 H-Pyrazole-1-carbothioamide derivatives as EGFR kinase inhibitors. ACS
omega. 2020; 5(30):18662-74.
8. Hendrickx R, Lamm Bergström E, Janzén DL, Fridén M, Eriksson U, Grime K, Ferguson
D. Translational model to predict pulmonary pharmacokinetics and efficacy in man for
inhaled bronchodilators. CPT: pharmacometrics & systems pharmacology. 2018;
7(3):147-57.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
1. Abdulrahman HL, Uzairu A, Uba S. Computer modeling of some anti-breast cancer
compounds. Structural Chemistry. 2021; 32(2):679-87.
2. Aykul S, Martinez-Hackert E. Determination of half-maximal inhibitory concentration
using biosensor-based protein interaction analysis. Analytical biochemistry. 2016;
508:97-103.
3. Changyong FE, Hongyue WA, Naiji LU, Tian CH, Hua HE, Ying LU. Log-
transformation and its implications for data analysis. Shanghai archives of psychiatry.
2014; 26(2):105.
4. da Silva Costa J, da Silva Lopes Costa K, Cruz JV, da Silva Ramos R, Silva LB, Do
Socorro Barros Brasil D, de Paula da Silva CH, Dos Santos CB, da Cruz Macedo WJ.
Virtual screening and statistical analysis in the design of new caffeine analogues
molecules with potential epithelial anticancer activity. Current pharmaceutical design.
2018; 24(5):576-94.
5. Ferreira EF, Silva LB, Costa GV, Costa JS, Fujishima MA, Leão RP, Ferreira AL,
Federico LB, Silva CH, Rosa JM, Macêdo WJ. Identification of new inhibitors with
potential antitumor activity from polypeptide structures via hierarchical virtual screening.
Molecules. 2019; 24(16):2943.
6. Hadaji EG, Bourass M, Ouammou A, Bouachrine M. 3D-QSAR models to predict anti-
cancer activity on a series of protein P38 MAP kinase inhibitors. Journal of Taibah
University for Science. 2017; 11(3):392-407.
7. Hajalsiddig TT, Osman AB, Saeed AE. 2D-QSAR modeling and molecular docking
studies on 1 H-Pyrazole-1-carbothioamide derivatives as EGFR kinase inhibitors. ACS
omega. 2020; 5(30):18662-74.
8. Hendrickx R, Lamm Bergström E, Janzén DL, Fridén M, Eriksson U, Grime K, Ferguson
D. Translational model to predict pulmonary pharmacokinetics and efficacy in man for
inhaled bronchodilators. CPT: pharmacometrics & systems pharmacology. 2018;
7(3):147-57.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
9. Kaushik P, Shakil NA, Rana VS. Synthesis, Biological Evaluation, and QSAR Studies of
3-Iodochromone Derivatives as Potential Fungicides. Frontiers in chemistry. 2021;
9:636882.
10. Liang T, Ren L, Li Q. 3D-QSAR studies of tetraoxanes derivatives as antimalarial agents
using CoMFA and CoMSIA approaches. Bulletin of the Korean Chemical Society. 2013;
34(6):1823-8.
11. Obadawo BS, Asogwa U, Ali AA. QSAR studies of BBR analogues against
coxsackievirus B1. Bulletin of the National Research Centre. 2022; 46(1):1-8.
12. Puzyn T, Gajewicz A, Leszczynska D, Leszczynski J. Nanomaterials–the next great
challenge for QSAR modelers. InRecent advances in QSAR studies 2010 (pp. 383-409).
Springer, Dordrecht.
13. Ramalakshmi N, Manimegalai P, Bhandare RR, Kumar SA, Shaik AB. 2D-Quantitative
structure activity relationship (QSAR) modeling, docking studies, synthesis and in-vitro
evaluation of 1, 3, 4-thiadiazole tethered coumarin derivatives as antiproliferative agents.
Journal of Saudi Chemical Society. 2021; 25(7):101279.
14. Roy K. Advances in QSAR modeling. Applications in Pharmaceutical, Chemical, Food,
Agricultural and Environmental Sciences; Springer: Cham, Switzerland. 2017; 555:39.
15. Sahu NK, Sharma MC, Mourya V, Kohli DV. QSAR studies of some side chain modified
7-chloro-4-aminoquinolines as antimalarial agents. Arabian Journal of Chemistry. 2014;
7(5):701-7.
16. Thakur A, Parashar A, Sharma VK, Kumar A, Mehta V. 2D-QSAR and Molecular
Docking based virtual screening of the herbal molecules against Alzheimers Disorder: An
approach to predict CNS activity. bioRxiv. 2022.
17. Vyas VK, Ghate M, Goel A. Pharmacophore modeling, virtual screening, docking and in
silico ADMET analysis of protein kinase B (PKB β) inhibitors. Journal of Molecular
Graphics and Modelling. 2013; 42:17-25.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
3-Iodochromone Derivatives as Potential Fungicides. Frontiers in chemistry. 2021;
9:636882.
10. Liang T, Ren L, Li Q. 3D-QSAR studies of tetraoxanes derivatives as antimalarial agents
using CoMFA and CoMSIA approaches. Bulletin of the Korean Chemical Society. 2013;
34(6):1823-8.
11. Obadawo BS, Asogwa U, Ali AA. QSAR studies of BBR analogues against
coxsackievirus B1. Bulletin of the National Research Centre. 2022; 46(1):1-8.
12. Puzyn T, Gajewicz A, Leszczynska D, Leszczynski J. Nanomaterials–the next great
challenge for QSAR modelers. InRecent advances in QSAR studies 2010 (pp. 383-409).
Springer, Dordrecht.
13. Ramalakshmi N, Manimegalai P, Bhandare RR, Kumar SA, Shaik AB. 2D-Quantitative
structure activity relationship (QSAR) modeling, docking studies, synthesis and in-vitro
evaluation of 1, 3, 4-thiadiazole tethered coumarin derivatives as antiproliferative agents.
Journal of Saudi Chemical Society. 2021; 25(7):101279.
14. Roy K. Advances in QSAR modeling. Applications in Pharmaceutical, Chemical, Food,
Agricultural and Environmental Sciences; Springer: Cham, Switzerland. 2017; 555:39.
15. Sahu NK, Sharma MC, Mourya V, Kohli DV. QSAR studies of some side chain modified
7-chloro-4-aminoquinolines as antimalarial agents. Arabian Journal of Chemistry. 2014;
7(5):701-7.
16. Thakur A, Parashar A, Sharma VK, Kumar A, Mehta V. 2D-QSAR and Molecular
Docking based virtual screening of the herbal molecules against Alzheimers Disorder: An
approach to predict CNS activity. bioRxiv. 2022.
17. Vyas VK, Ghate M, Goel A. Pharmacophore modeling, virtual screening, docking and in
silico ADMET analysis of protein kinase B (PKB β) inhibitors. Journal of Molecular
Graphics and Modelling. 2013; 42:17-25.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted October 18, 2022.;https://doi.org/10.1101/2022.10.15.512366doi:bioRxiv preprint
100%
