Page 1 of 17
British Journal of Healthcare and Medical Research - Vol. 10, No. 2
Publication Date: April 25, 2023
DOI:10.14738/jbemi.102.14405.
Saxena, S., Kumar, M., Saxena, S. N., Prakash, P. S., Agarwal, D., Bhutia, D. D., & Vishal, M. K. (2023). Machine Learning Modelling of
Effective Cumin (Cuminnum cyminum L.) Genotype for Antipyretic Property Based on Secondary Metabolites. British Journal of
Healthcare and Medical Research, Vol - 10(2). 311-327.
Services for Science and Education – United Kingdom
Machine Learning Modelling of Effective Cumin (Cuminnum
cyminum L.) Genotype for Antipyretic Property Based on
Secondary Metabolites
Shubhi Saxena
Department of Biomedical Engineering,
School of Engineering and Technology,
Mody University, Lakshmangarh, Sikar,
Rajasthan- 332311, India
Manish Kumar
Department of Biomedical Engineering,
School of Engineering and Technology,
Mody University, Lakshmangarh, Sikar,
Rajasthan- 332311, India
S N Saxena
ICAR-National Research Centre on Seed Spices,
Tabiji, Ajmer-305206, India
P Shakti Prakash
Department of Biomedical Engineering,
School of Engineering and Technology,
Mody University, Lakshmangarh, Sikar,
Rajasthan- 332311, India
Dolly Agarwal
ICAR-National Research Centre on Seed Spices,
Tabiji, Ajmer-305206, India
Dawa Doma Bhutia
Department of Biomedical Engineering,
School of Engineering and Technology,
Mody University, Lakshmangarh, Sikar,
Rajasthan- 332311, India
Mukesh Kumar Vishal
ICAR-National Research Centre on Seed Spices,
Tabiji, Ajmer-305206, India and Indian Institute
of Technology Bombay, Powai, Mumbai-400076
Page 2 of 17
312
British Journal of Healthcare and Medical Research (BJHMR) Vol 10, Issue 2, April- 2023
Services for Science and Education – United Kingdom
ABSTRACT
Cumin is a flowering plant that has been used as a spice in multicuisine. It has
various health benefits such as weight loss, cholesterol control, anti-diabetes, and
more. It also consists of dietary fibers, vitamin B, vitamin E, and minerals, especially
iron and magnesium. The present study was conducted for the effective
implementation of two tasks, the prediction of cumin genotype in which the inputs
were total phenolic, flavonoid, and antioxidant content and the classification of
anti-pyretic activity. MLP and other ML algorithm simulations were committed to
executing the tasks of prediction and classification. The effectiveness of the
proposed model was compared with the various classification and regression
techniques like SVM, Naïve Bayes, KNN, Logistic Regression, and Decision Tree.
Along with the mentioned task, this paper also exhibits the implementation of
feature selection techniques like PCA in ML-based prediction and classification. It
was found that MLP with PCA has outperformed other algorithms.
Keywords: Cumin (Cuminnum cyminum L.), Forecasting, Genotypes, Machine Learning,
Secondary Metabolites
INTRODUCTION
Cumin is an angiosperm that belongs to the family Apiaceae, i.e., native to Egypt but cultivated
widely in India and middle Asia for its aromatic seeds and its oil used for medicinal purposes.
It is widely used in Indian, German, Middle Eastern, Italian, and Mexican cooking as spice [1]
.
The economic part of the cumin plant is its seeds are used dry both as a whole and as ground
form, Figure 1-a. It is produced in areas with arid and semi-arid climatic conditions, which are
found in most Asian countries like India, China, Iran, and Indonesia. However, India is the
largest producer of cumin which is 70% of the entire world's production. The main cumin
growing belt in India includes the state of Rajasthan and Gujrat[2]–[4]
, these areas are deinetaed
in the Figure 1-b. It is very important cash crop for the farmers in the arid and semi-arid regions
of the India, and requires low relative humidity specially at the time of maturity. Cumin has
various advantages in health care applications as it helps in improving digestion, boosting
immunity, and preventing the growth of cancerous cells. Due to its good anti-inflammatory as
well as antipyretic properties, it helps in reducing pain and fever. Also, it is a good source of
antioxidants [5]
.
In our earlier study, widespread cumin genotypes were collected from different cultivating
areas of India and evaluated for the presence of phenolics, flavonoids, antioxidants [6],[7] and
fatty acids, as well as their essential oil content and constituents like cumin aldehyde, gamma- terpinene, alpha-pinene, etc. [8], [9]
. It also evaluated the effectiveness of different solvent
extracts of cumin seeds and leaves on the extraction of phenol, flavonoids, and antioxidant
contents. Maximum phenolic contents were found in the methanol extracts irrespective of
cumin genotype, and a minimum of total phenolic contents was observed in genotype RZ-19.
The experimental evaluation of the genotypic variation for the therapeutic use of cumin been
carried out by Agrawal et al. [2]
. In their work, the cumin seeds of different genotypes i.e., GC-1,
GC-4, RZ-19, RZ-209, RZ-341 were extracted using hexane and methanol as solvent. Further,
these crude seed extracts were used as a drug to study various medicinal activities such as
antipyretic properties, anti-diabetic, anti-fungal and glucose tolerance test [10]–[13][12], [13]. The in
vivo observation revealed that the pure genotype of cumin was found to have more additional
therapeutic properties.
Page 3 of 17
313
Saxena, S., Kumar, M., Saxena, S. N., Prakash, P. S., Agarwal, D., Bhutia, D. D., & Vishal, M. K. (2023). Machine Learning Modelling of Effective Cumin
(Cuminnum cyminum L.) Genotype for Antipyretic Property Based on Secondary Metabolites. British Journal of Healthcare and Medical Research,
Vol - 10(2). 311-327.
URL: http://dx.doi.org/10.14738/jbemi.102.14405.
Nowadays, Machine Learning algorithms are increasingly important in biology due to their
precision, accuracy, and robustness as they capture the complexity and pattern of small to large
datasets [20]. Therefore, in the present study, a model was designed using Artificial Neural
Network (ANN), taking the data set based on the results obtained from the above-mentioned
work. Using this model, a simulation environment was created to classify and predict the
effective genotypes of cumin based on secondary metabolites. The ANN belongs to the family of
Machine Learning (ML), which mimics the human brain and how it performs every other
function [21]. It uses a learning model that can perform adjustments independently. In return,
ML is a part of Artificial Intelligence (AI) [27]-[30]
, and it functions to allow software applications
to become more accurate at predicting the outcomes without being formally programmed to do
so. ANN uses a learning model that can perform adjustments independently. This tool is very
useful for non-linear statistical data modelling [14]
. The different experiments and modelling
have been performed using ANN, including prediction and classification. Despite initial
research, the genotypic and phenotypic characteristic prediction still requires more attention
[15]–[17]
. The contribution of current work is mentioned as, in this work, two major tasks were
performed with the help of the ANN model. The first one is the prediction of genotype and the
second one is the classification of the antipyretic activity.
1. The developed technique will provide a parallel simulation environment that allows the
researchers to predict not only genotype but also classify any therapeutic property of
cumin.
2. With the use of such technique in performing any type of experiment, the process will
become easy and less time-consuming and any unknown data can be tested using the
existing data set as it will provide some level of probability and accuracy in the results.
This work, will strive to get a result that is satisfying and may be able to reduce the load of
performing the experiments such that researchers who would want to further experiment on
cumin can replace this part of the experimental phase with the help of our model.
MATERIAL AND METHODS
Data Set Formation and Algorithm Implementation
The data set was acquired from the experiment performed by Aggarwal, et. al. [2]
. The author
used the crude seed extract of cumin as a drug after extracting by hexane and methanol
solvents. The experiment was performed on the albino mice by injecting the cumin seed crude
extracts. The mice were first injected with carrageenan, which led to an increase in their body
temperature (fever). To control their body temperature, the tone of the mice was given with
allopathic medicine as control and others were given methanol and hexane crude seed extracts
of different cumin genotypes as experimental drugs. The observations were taken at the time
interval of 2 hours and were checked which among the extracts were able to bring the body
temperature of the mice equal to or similar to the body temperature of the control mice. Most
of the genotype extracts were able to control the body temperature of the mice but a little less
than the control mice, it was found that genotype GC-1 of methanol extract was able to reach
the level of control mice and worked effectively on it. Using this experiment and with the help
of ML, this model was functioned for the classification of antipyretic properties based on the
effectiveness of each genotype extract on defined time interval and prediction of genotype
based on the different essential oil contents of all given genotype. For the prediction of
genotype, total phenolic flavonoid, and anti-oxidant data present in the crude seed extract of
cumin genotype were used and for the classification purpose, the data on how does the different