SYSTEMATIC REVIEW

Artificial Intelligence–Driven Discovery of Plant-Derived Bioactive Compounds in Cancer Therapeutics: Computational Approaches and Molecular Target Identification

Irshad Khan¹ | Asad Khalid² | Muitaba Ehsan² | Muhammad Akram¹ | Ejaz Ehsan Khan^3* | Khaleeq Ullah¹

¹Department of Pathology, University of Lahore, Lahore, Pakistan | ²Department of Biosciences, COMSATS University, Islamabad, Pakistan
³Department of Pathology, University of Florence, Italy
*Correspondence: Ejaz Ehsan Khan (ejazahmedpatho@gmail.com)

Received: 04 May, 2026; Revised (1^st): 08 June, 2026; Revised (2^nd): 08 September, 2026; Accepted: 12 September, 2026; Published: 20 September, 2026

INTRODUCTION

Cancer was among the leading causes of morbidity and mortality, with an estimated 19.3 million new cases and about 10 million deaths reported in 2020 alone ¹. Even with significant improvements in surgery, radiotherapy, and conventional chemotherapy, plant-derived bioactive compounds offer more effective, safe, and specific anticancer agents, which provide multi-targeted pharmacological properties and also overcome the concerns related to drug resistance ^2,3. Nevertheless, the traditional experimental process of discovery and confirmation of such compounds was laborious, time-consuming, and resource-intensive ⁴.

The advent of machine learning (ML) and artificial intelligence (AI) has revolutionized the drug discovery and development field ⁵. Computational methods, such as deep learning, quantitative structure-activity relationship (QSAR) modelling, network pharmacology, molecular docking and molecular dynamics (MD) simulation, can screen tens of thousands of phytochemicals with respect to molecular targets and map multi-target pharmacological networks at an unprecedented speed and accuracy ⁶. The built-in databases like the Traditional Chinese Medicine Systems Pharmacology Database (TCMSP), PubChem, DrugBank, ChEMBL, and SwissTargetPrediction also assist AI-based virtual pipelines with curated chemical, biological, and clinical data ⁷.

Although the AI-assisted phytochemical studies were increasing rapidly, a comprehensive synthesis consolidating the evidence on the AI-based computational strategies to identify plant-derived anticancer bioactives, to characterize their molecular targets, and validate their interactions by in silico and in experimental systems, was still lacking ⁸. The main objective of the systematic review was to summarize the data from computational pharmacology studies, including network pharmacology, molecular docking, molecular dynamics simulations, and AI/ML, on the identification and validation of plant-derived bioactive compounds with anticancer properties to assess the range of computational methods used, index the databases and virtual screening pipelines, examine the applicability of multi-target network pharmacology to explain polypharmacological mechanisms, and critically evaluate the limitations and translation issues of in silico predictions in phytochemical oncology.

METHODOLOGY

Due to the direct retrospective design of this systematic review, which relies entirely on previously published, aggregated public data, this study was not prospectively registered in an international repository. However, to guarantee scientific rigor, the research layout, search strings, and evaluation criteria were mapped internally and strictly followed the structural framework of the PRISMA 2020 statement ⁹.

Database Search: An extensive search of electronic literature was performed in four scientific databases: PubMed, Scopus, Web of Science, and Google Scholar to find publications published from October 2020 to July 2025.

Search Strategy Used: The search strategy was developed as a structured search with the help of MeSH terms and free-text keywords, along with the Boolean operators (AND, OR). The main keywords were: "artificial intelligence ", "machine learning ", "deep learning ", "QSAR ", "network pharmacology ", "molecular docking ", "phytochemicals ", "plant-derived compounds ", "neoplasm ", "molecular target ", "protein-ligand interaction ", "signaling pathway ", "in silico ", and "computational pharmacology ". Other records were identified during manual review of reference lists of the qualified full-text articles.

Inclusion and Exclusion Criteria: The studies included were original research articles, used at least one AI or computational method, studied plant-derived phytochemicals, investigated anticancer effects or molecular targets in cancer, used known computational databases, and were published in English and provided free access to the full text. The studies were excluded if they were review article, conference abstracts, editors, or book chapters.

Study Selection Process: All database search results were imported into a reference management system, and duplicates were removed. All remaining records were then screened by two independent reviewers (ME, MA) in two distinct stages: title and abstract screening and got same same results. Full-text screening of the remaining articles was done. Any disagreement between reviewers at any level was settled by discussion.

Data Extraction: The data were independently extracted by two reviewers (ME, MA), comprising study design, AI and computational methods used, database and bioinformatics tools used, source of plant and specific bioactive compound(s), type(s) of cancer investigated, molecular target, identified signaling pathways, validation methodology, and key findings. Any discrepancies among reviewers were resolved through discussion.

Primary Outcome: The main focus of the systematic review was to describe and review the AI/computational methods of identifying and verifying plant-derived bioactive compounds as anticancer agents with particular attention to the molecular targets and signaling pathways.

Quality Assessment: There is no validated method for in silico studies, the risk of bias was assessed using an adapted version of the PROBAST. PROBAST domains were organized as follows: participants (data sources/databases), predictors (compounds/features), outcomes (molecular targets/pathways), and analysis (computational methods such as docking, network pharmacology, and machine learning). For studies involving experimental validation, appropriate domains from the Joanna Briggs Institute Critical Appraisal Tools were also evaluated. The domains were graded as low (+), high (−), or unclear (?). Studies with ≥1 high-risk domain were categorized as overall high risk, with focus on the analysis domain and validation status ¹⁰. The risk of bias was assessed with the SYRCLE tool for in vitro and in vivo research requirements. The certainty of evidence was evaluated according to the GRADE approach.

Figure 1: A PRISMA flow diagram following PRISMA guidelines 2020 outlining the study selection process of the systematic review

RESULTS

The first search of electronic databases retrieved 342 records (PubMed: 128, Scopus: 94, Web of Science: 76, Google Scholar: 44). Following automated and manual elimination of duplicate records, 219 unique articles were left. An additional 98 records failed to pass title and abstract screening and were excluded due to non-conformity with inclusion criteria, mainly because they were review articles, did not deal with plant-derived compounds or did not contain any in silico computational aspect. The 121 full-text articles were evaluated, 103 were eliminated due to the following reasons: not available freely in the full-text (n = 35), not a primary AI/computational component (n = 22), review or non-original research design (n = 19), only non-cancer diseases (n = 14), not in English (n = 7), and lack of quantitative computational outcomes. Eventually, 18 studies were included in this systematic review and met all the eligibility criteria. The PRISMA flow diagram in Figure 1, summarize the process of study selection.

The main findings of all 18 studies incorporated into this systematic review were given in Table I, and all of them utilized at least one of the AI-driven or computational pharmacology methods to explore the use of plant-derived bioactive compounds in anticancer applications. Studies included in this review contain network pharmacology, molecular docking, molecular dynamics simulation, QSAR modelling, machine learning, and pharmacophore-based approaches representing a variety of cancer types. Different bioinformatics databases were used to find phytochemical targets and map pertinent disease pathways.

Table I: Summary Table Showing Characteristics and Key Findings of Studies Included in the Systematic Review

The 18 studies addressed a variety of phytochemicals of medicinal plants and a variety of cancers. The most widely used computational model was network pharmacology along with molecular docking, followed by machine learning to train predictive anticancer activity models. The most frequently identified bioactive phytochemical was quercetin, which was used against breast cancer, hepatocellular carcinoma, and lung cancer via regulation of AKT1, EGFR, TP53, and MAPK signaling pathways. Among the included studies, 10 were purely based on in silico findings and 8 were supported by some experimental validations i.e., 8 studies used in vitro cytotoxicity measurements, with 2 studies also utilizing in vivo animal models, providing multi-level mechanistic confidence to the computational findings.

The study designs ranged from pure computational pharmacology to integrative in silico, in vitro, and in vivo platforms, indicating the steadily advanced AI-aided phytochemical research. The molecular targets that were observed in multiple studies were largely biased by a small group of high-connectivity oncoproteins, in particular, EGFR, AKT1, TP53, ESR1 and MAPK3, which are repeated across cancer types and plant sources, implying that the hub targets of phytochemical intervention are the intersections. KEGG pathway enrichment analysis demonstrated that PI3K/ Akt, MAPK, NF-kB cascades were consistently enriched by plant-derived compounds. Notably, results from all studies that included in vitro or in vivo validation were consistent with in silico predictions, which indicated biological plausibility of the computational methods used.

The PROBAST-based assessment showed that the majority of research were at high risk of bias, resulting from limits in the analysis domain, such as a lack of external validation and reliance on computational predictions. Studies that included in vitro and/or in vivo validation had lower risk in outcome areas, but pure in-silico studies were consistently assessed as high risk. Applicability problems were typically minimal, while several studies had unclear applicability due to inadequate reporting of outcome definitions and predictor selection, as shown in Table II.

Table II: Risk of bias assessment of the included studies using adapted PROBAST (Prediction Model Risk of Bias Assessment Tool)

The risk of bias assessment of the eight studies that involved in vitro experimental validation showed a similar tendency of some concerns in all the studies, as shown in Table III. The area of random sequence generation was homogeneously considered as Low risk because all the studies used randomized layouts and indicated cell line sources and cell culture conditions. The allocation concealment and result assessment are rated as some concerns in all eight studies, which is a well-known weakness of in vitro laboratory-based research. The unfinished outcome data and selective reporting areas were considered low risk in all situations because studies reported complete data of all biological replicates and assessed outcomes.

Table III: Risk of Bias Assessment of In Vitro Studies using Modified SYRCLE Tool

The risk of bias assessment of the two studies that used in vivo animal model validation has produced an overall rating of some concerns, as shown in Table IV. The process of random sequence generation was also low risk in both studies, with animals being randomly assigned to experimental groups with well-defined group sizes. Generally, the in vivo experiments showed that preclinical animal research has acceptable methodological standards, but some concerns rating indicated that it would be better to be clearer about how blinding and allocation procedures were carried out in subsequent in vivo phytochemical validation experiments instead of showing that the reported results were systematically biased.

Table IV: Risk of Bias assessment of In Vivo Studies using SYRCLE Tool

Traffic light plot showing the risk of bias of the two in vivo preclinical animal studies were assessed using Robvis software, as shown in Figure 2. The small sample size of in vivo studies (n = 2) limited the generalizable conclusions on the quality of the overall preclinical evidence but the findings indicated that subsequent in vivo validation studies should incorporate and explicitly describe standardized blinding and randomization protocols to enhance the translational plausibility of computational predictions.

Figure 2: Traffic light plot illustrating the risk of bias assessment for in vivo preclinical animal studies using the Robvis tool

Traffic light plot illustrating the risk of bias measurement of the eight in vitro experimental studies were assessed with the Robvis software, as shown in Figure 3. Low risk of bias denoted by green circles and some concerns was denoted by yellow circles The plot indicated a consistent trend of low risk in random sequence generation, incomplete outcome reporting, and selective reporting, whereas the domain of allocation concealment and blinding always had some concerns. The certainty of evidence based on a GRADE framework was low, mainly because of the sheer indirectness of computational predictions based on the clinical outcome, methodological heterogeneity across docking platforms, and the limited number of experimental validations of the included studies.

Figure 3: Traffic light plot depicting the risk of bias assessment for in vitro experimental studies using Robvis toll

Overall, the studies demonstrated substantial variation in computational approaches, phytochemical sources, cancer types, molecular targets, and validation methods. The following discussion interprets these findings in the context of current evidence on computational phytochemical discovery and anticancer mechanisms

DISCUSSION

This systematic review summarized 18 original research studies which used artificial intelligence and computational pharmacology models such as network pharmacology, molecular docking, molecular dynamics simulation, QSAR modelling and machine learning to identify and predict the anticancer properties of plant-derived bioactive compounds. This integrative strategy demonstrated a reproducible convergence strategy in phytochemical targets and signaling pathways that would remain invisible in single-compound or single-cancer studies, providing a more general and transferable framework for guiding future in silico phytochemical oncology pipelines. Together, these methods have evolved into the backbone of modern phytochemical drug discovery, allowing large libraries of compounds to be screened simultaneously against a variety of cancer-relevant targets, and accelerating the process of identifying leads for experimental validation.

Among the 18 included studies, the most prevalent method was network pharmacology, followed by molecular docking. The frequent emergence of AKT1, EGFR, TP53, ESR1, and MAPK3 as central targets for phytochemicals derived from a broad range of plants and interrogating multiple cancers was biologically significant ²⁹. The PI3K/Akt/mTOR pathway was one of the most frequently overactive survival pathways in human cancer, and has been widely documented to be regulated by natural products ³⁰.

The most frequently identified bioactive compound, quercetin was identified in studies targeting breast cancer, hepatocellular carcinoma, and lung cancer with inhibition of AKT1 and EGFR, suppression of MAPK, and induction of apoptosis. This multifaceted activity was consistent with its reported modulation of PI3K/Akt, NF-κB and Notch pathways in a variety of tumors ³¹. Resveratrol has displayed anticancer activities, particularly in endometrial cancer and HCC, where it targeted TP53, AKT1, PIK3CA, CASP3, and EGFR via the PI3K/Akt and MAPK pathways ³². Resveratrol has shown cytoprotective effects through scavenging of reactive oxygen species, further justifying its inclusion in the virtual screening cycle for designing anti-cancer drugs ³³. The multiple hits on different oncogenic targets by structurally diverse phytochemicals support their polypharmacological activity and multi-target mechanisms, extending beyond conventional single-target approaches ³⁴.

The epidermal growth factor receptor (EGFR) emerged as the most prominent molecular target, Featuring in studies of breast, lung, liver, colorectal and cervical cancers. EGFR, commonly mutated in non-small cell lung cancer (NSCLC), has been extensively targeted by established therapies and continues to be explored through emerging immunotherapeutic strategies ³⁵. Plant-based EGFR inhibitors identified by high-throughput virtual screening and machine learning-driven QSAR studies were an opportune approach to address this clinical challenge.

The use of machine learning in two studies expanded phytochemical discovery beyond traditional network pharmacology approaches, by allowing large-scale activity prediction from molecular fingerprints using Random Forest, Support Vector Machine (SVM), k-Nearest Neighbour (kNN) and XGBoost models. This enabled the discovery of candidate molecules from the hundreds of thousands of natural product scaffolds present in the databases with a computational speed that could not be achieved with experimental screening. The in vivo study that validated the new candidate compounds and the experimental evidence that plant-derived withanolide E from Physalis peruviana exhibited anticancer activity in DMBA-induced breast cancer models ^{36, 37}, together strengthen the growing evidence that phytochemicals prioritized by computational approaches can have preclinical activity.

The results from studies related to hepatocellular carcinoma (HCC) focused on phytochemicals for the treatment of HCC, with network pharmacology analyses pinpointing RELA, SQSTM1, AKT1, EGFR, and TNF as potential targets. Natural products have demonstrated significant potential in targeting autophagic signaling pathways in HCC ³⁸, and the active compounds identified in the included studies, such as curcumol, quercetin, and Solanum surattense compounds, regulated PI3K/Akt, MAPK and NF-κB pathways that were strongly associated with autophagic and apoptotic processes in liver cancer ³⁹. The in vitro validation studies demonstrated target gene down-regulation in human liver cancer cell lines, thereby validating the computational findings, and highlighting the significance of network pharmacology-predicted targets in HCC.

However, critical translational gaps remain, and the PROBAST-based risk of bias assessment revealed that the majority of included studies were at high risk of bias, due to limitations in the analysis domain, specifically a lack of external validation and reliance solely on computational predictions without experimental confirmation. Only one study had a low overall risk of bias, due to the inclusion of both in vitro and in vivo validation. Pure in silico investigations were consistently evaluated as high-risk, emphasizing the importance of prospective experimental validation. Pharmacokinetic filters used for screening (oral bioavailability ≥ 30%, Lipinski's Rule of Five) don't completely predict in vivo properties. Phytochemicals like quercetin, curcumin, and resveratrol were heavily metabolized, have poor systemic bioavailability, and insufficient penetration into the tumor, limiting the amount of active drug that can reach the target site.

Limitations of this review are process-related (e.g., confined to English-language, full-text articles published in 2020-2025 in four databases), which can lead to language and publication bias. Also, there was certain subjectivity inherent in the adapted application of the PROBAST and SYRCLE tools to assess the risk of bias, and the exclusion of non-peer-reviewed literature (e.g., conference abstracts) could have left some developing results out. There was a risk of publication bias towards positive docking results, which increases the apparent consistency in silico agreement. Future research should focus on hybrid validation strategies (integrating in vitro and in vivo assays from the outset) , adopt standardized reporting guidelines for computer-aided drug design, and use of nano-formulation strategies to enhance the translational potential of in silico-prioritized phytochemicals.

CONCLUSION

Computational pharmacology techniques (including network pharmacology, molecular docking, molecular dynamics simulation, and machine learning) have become essential to discovering plant-derived leads for cancer treatment. Across 18 original studies, a wide range of phytochemicals were shown to engage high -priority promising targets (AKT1, EGFR, TP53, ESR1) through signaling pathways (PI3K/Akt, MAPK, NF-κB). Although the overall certainty of evidence was very low because of the indirect nature of the computational models, the plausibility of biological findings was supported by consistent experimental verification in some of the studies. Transparent reporting, prospective in vitro/in vivo validation, and drug delivery approaches (nano-formulations) are critical steps that are needed to translate these computational findings into anticancer leads for future research.

References

1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71(3):209–249. https://doi.org/10.3322/caac.21660
2. Newman DJ, Cragg GM. Natural Products as Sources of New Drugs over the Nearly Four Decades from 01/1981 to 09/2019. J Nat Prod. 2020;83(3):770–803. https://dx.doi.org/10.1021/acs.jnatprod.9b01285
3. Asma ST, Acaroz U, Imre K, Morar A, Shah SR, Hussain SZ, et al. Natural products/bioactive compounds as a source of anticancer drugs. Cancers. 2022;14(24):6203. https://doi.org/10.3390/cancers14246203
4. Kopustinskiene DM, Jakstas V, Savickas A, Bernatoniene J. Flavonoids as anticancer agents. Nutrients. 2020;12(2):457. https://doi.org/10.3390/nu12020457
5. Sadybekov AV, Katritch V. Computational approaches streamlining drug discovery. Nature. 2023;616(7958):673–685. https://doi.org/10.1038/s41586-023-05905-z
6. Tropsha A, Isayev O, Varnek A, Schneider G, Cherkasov A. Integrating QSAR modelling and deep learning in drug discovery: the emergence of deep QSAR. Nat Rev Drug Discov. 2024;23(2):141-55. https://doi.org/10.1038/s41573-023-00832-0
7. Ru J, Li P, Wang J, Zhou W, Li B, Huang C, et al. TCMSP: a database of systems pharmacology for drug discovery from herbal medicines. J Cheminformatics. 2014;6(1):13. https://doi.org/10.1186/1758-2946-6-13
8. Nogales C, Mamdouh ZM, List M, Kiel C, Casas AI, Schmidt HH. Network pharmacology: curing causal mechanisms instead of treating symptoms. Trends Pharmacol Sci. 2022;43(2):136-150. https://doi.org/10.1016/j.tips.2021.11.004
9. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372. https://doi.org/10.1136/bmj.n71
10. Gao J, Yao Y, Xue J, Chen R, Yang X, Xu J, et al. Methodological conduct and risk of bias in studies on prenatal birthweight prediction models using machine learning techniques: a systematic review. BMC Pregnancy and Childbirth. 2025 Jul 2;25(1):696. https://doi.org/10.1186/s12884-025-07727-5
11. Sakle NS, More SA, Mokale SN. A network pharmacology-based approach to explore potential targets of Caesalpinia pulcherima: an updated prototype in drug discovery. Sci Rep. 2020;10(1):17217. https://doi.org/10.1038/s41598-020-74251-1
12. Zhao F, Qin J, Liang Y, Zhou R. Exploring anti-liver cancer targets and mechanisms of oxyresveratrol: in silico and verified findings. Bioengineered. 2021;12(2):9939–9948. https://doi.org/10.1080/21655979.2021.1985328
13. Zhong Z, Guo X, Zheng Y. Network pharmacology-based and molecular docking analysis of resveratrol's pharmacological effects on type I endometrial cancer. Anticancer Agents Med Chem. 2022;22(10):1933–1944. https://doi.org/10.2174/1871520621666211015140455
14. Shaji SK, Drishya G, Sunilkumar D, Suravajhala P, Kumar GB, Nair BG. Systematic understanding of anti-tumor mechanisms of tamarixetin through network and experimental analyses. Sci Rep. 2022;12(1):3966. https://doi.org/10.1038/s41598-022-07087-6
15. Huang X, Rehman HM, Szöllősi AG, Zhou S. Network pharmacology-based approach combined with bioinformatic analytics to elucidate the potential of curcumol against hepatocellular carcinoma. Genes. 2022;13(4):653. https://doi.org/10.3390/genes13040653
16. Khalid HR, Aamir M, Tabassum S, Alghamdi YS, Alzamami A, Ashfaq UA. Integrated system pharmacology approaches to elucidate multi-target mechanism of Solanum surattense against hepatocellular carcinoma. Molecules. 2022;27(19):6220. https://doi.org/10.3390/molecules27196220
17. Wu Z, Kang J, Tan W, Wei C, He L, Jiang X, et al. In silico and in vitro studies on the mechanisms of Chinese medicine formula (Yiqi Jianpi Jiedu formula) in the treatment of hepatocellular carcinoma. Comput Math Methods Med. 2022;2022(1):8669993. https://doi.org/10.1155/2022/8669993
18. Bultum LE, Tolossa GB, Kim G, Kwon O, Lee D. In silico activity and ADMET profiling of phytochemicals from Ethiopian indigenous aloes using pharmacophore models. Sci Rep. 2022;12(1):22221. https://doi.org/10.1038/s41598-022-26446-x
19. Bhattacharya K, Nath BC, Ahmed E, Khanal P, Chanu NR, Deka S, et al. Integration of network pharmacology, molecular docking, and simulations to evaluate phytochemicals from Phyllanthus sp. against cervical cancer. RSC Adv. 2024;14(6):4188–4200. https://doi.org/10.1039/d3ra06297j
20. Arif R, Bukhari SA, Mustafa G, Ahmed S, Albeshr MF. Network pharmacology and experimental validation to explore the potential mechanism of Moringa oleifera for the treatment of breast cancer. Pharmaceuticals (Basel). 2024;17(5):617. https://doi.org/10.3390/ph17050617
21. Varadharajan V, Balu AK, Shiju A, Muthuramalingam P, Shin H, Venkidasamy B, et al. Deciphering the anticancer arsenal of Piper longum: network pharmacology and molecular docking unveil phytochemical targets against lung cancer. Int J Med Sci. 2024;21(10):1915. https://doi.org/10.7150/ijms.98393
22. Saini H, Gupta PK, Mahapatra AK, Rajagopala S, Tripathi R, Nesari T. Deciphering the multi-scale mechanism of herbal phytoconstituents in targeting breast cancer: a computational pharmacological perspective. Sci Rep. 2024;14(1):23795. https://doi.org/10.1038/s41598-024-75059-z
23. Irham LM, Adikusuma W, Illian DN, Mazaya M, Mubaraq A, Ma’ruf M, et al. Deciphering anti-colorectal cancer potential of Avicennia alba bioactives via network pharmacology and in vitro validation. Sci Rep. 2025;15(1):27976. https://doi.org/10.1038/s41598-025-12500-x
24. Gao K, Chen Z, Zhang N, Jiang P. High throughput virtual screening and validation of plant-based EGFR L858R kinase inhibitors against non-small cell lung cancer: An integrated approach utilizing GC–MS, network pharmacology, docking, and molecular dynamics. Saudi Pharm J. 2024;32(9):102139. https://doi.org/10.1016/j.jsps.2024.102139
25. Gondokesumo ME, Rasyak MR. In-silico prediction of anti-breast cancer activity of ginger (Zingiber officinale) using machine learning techniques. Breast Dis. 2024;43(1):99-110. https://doi.org/10.3233/BD-249002
26. Grijaldo SJ, Alvarez MR, Heralde III FM, Nacario RC, Lebrilla CB, Rabajante JF, et al. Integrating Computational Methods in Network Pharmacology and In Silico Screening to Uncover Multi-targeting Phytochemicals against Aberrant Protein Glycosylation in Lung Cancer. ACS Omega. 2023;8(26):23833-23844. https://doi.org/10.1021/acsomega.3c01825
27. Quintal Bojórquez ND, Morales Mendoza LF, Hidalgo-Figueroa S, Hernández Álvarez AJ, Segura Campos MR. In silico analysis of the interaction of de novo peptides derived from Salvia hispanica with anticancer targets: Evaluation of the anticancer potential of de novo peptides derived from Salvia hispanica through molecular docking. J Biomol Struct Dyn. 2024;42(12):6119-6135. https://doi.org/10.1080/07391102.2023.2232045
28. Das AP, Mathur P, Agarwal SM. Machine learning, molecular docking, and dynamics-based computational identification of potential inhibitors against lung cancer. ACS Omega. 2024;9(4):4528-39. https://doi.org/10.1021/acsomega.3c07338
29. Yang J, Gu J, Shen Y, Cao L, Zhou H, Zhu W. Effect of Shan Zha (Hawthorn or Crataegus) on gastrointestinal cancer: a network pharmacology and molecular docking study. Cancer Pathog Ther. 2023;1(4):229–237. https://doi.org/10.1016/j.cpt.2023.02.001
30. Morgos DT, Stefani C, Miricescu D, Greabu M, Stanciu S, Nica S, et al. Targeting PI3K/AKT/mTOR and MAPK signaling pathways in gastric cancer. Int J Mol Sci. 2024;25(3):1848. https://doi.org/10.3390/ijms25031848
31. Zenjanab MK, Hashemzadeh N, Alimohammadvand S, Sharifi-Azad M, Abdolahinia ED, Jahanban-Esfahlan R. Notch signaling suppression by golden phytochemicals: potential for cancer therapy. Adv Pharm Bull. 2024;14(2):302. https://doi.org/10.34172/apb.2024.035
32. Shahani SB, Shaikh P, Memon SG, Memon S. Possible prevention of reactive oxygen species induced human trabecular meshwork cell damage by resveratrol and ascorbic acid. Prof Med J. 2019;26(7):1036–41. https://doi.org/10.29309/TPMJ/2019.26.07.3229
33. Chen Y, Tan X, Zhang W, Li Y, Deng X, Zeng J, et al. Natural products targeting macroautophagy signaling in hepatocellular carcinoma therapy: recent evidence and perspectives. Phytother Res. 2024;38(3):1623–1650. https://doi.org/10.1002/ptr.8103
34. Khaliq M, Aftab K, Ikhlaq A, Ahmed A. Drug Repurposing Potential of Natural Products: Molecular Targets, Pharmacological Profiles, and Precision Therapeutic Strategies: A Narrative Review. J Biomol Pathog Ther. 2026;2(3):156-162. https://doi.org/10.68041/jbpt.v2i3/09
35. Radwan AA, Alanazi F, Al-Dhfyan A. Bioinformatics-driven discovery of novel EGFR kinase inhibitors as anti-cancer therapeutics: in silico screening and in vitro evaluation. PLoS One. 2024;19(4). https://doi.org/10.1371/journal.pone.0298326
36. Gul F, Memon S, Ujjan I, Goswami P, Bhatti KA. The effects of 4β-hydroxy withanolide E extracted from Physalis Peruviana on Complete Blood Count of Dimethylbenz(a)anthracene-induced Breast Cancer in Albino Rats. Adv Life Sci. 2023;10(3):434–8. https://doi.org/10.62940/als.v10i3.1861
37. Xie X, Gui L, Qiao B, Wang G, Huang S, Zhao Y, et al. Deep learning in template-free de novo biosynthetic pathway design of natural products. Brief Bioinform. 2024;25(6). https://doi.org/10.1093/bib/bbae495
38. Iweala EJ, Oluwapelumi AE, Dania OE, Ugbogu EA. Bioactive phytoconstituents and their therapeutic potentials in the treatment of haematological cancers: a review. Life (Basel). 2023;13(7):1422. https://doi.org/10.3390/life13071422
39. Dixit N, Motwani H, Patel SK, Rawal RM, Solanki HA. Decoding the mechanism of andrographolide to combat hepatocellular carcinoma: a network pharmacology integrated molecular docking and dynamics approach. J Biomol Struct Dyn. 2024;42(19):10237-55. https://doi.org/10.1080/07391102.2023.2256866

AL-NAJAM Journal of Medical and Life Sciences, Al Najam J Med Life Sci 2026:1(3), p83-92 (anjmls.org) © 2026 Authors. This work is published by Multidisciplinary Scholarly Advancement and Research MSAR Institute. The full terms of Journal Publishing policy is available at https://anjmls.com/index.php/anjmls/publishing-license and incorporate the Creative Commons Attribution – Non Commercial (CC BY, NC 4.0) License https://creativecommons.org/licenses/by-nc/4.0/. By accessing the work you hereby accept the Terms. Non-commercial uses of the work are permitted without any further permission, provided the work is properly attributed. Publisher’s Note: MSAR Institute remains neutral with regard to jurisdictional claims in published maps and institutional affiliations, and assumes no liability for the scientific accuracy or clinical efficacy of the content herein, as they rest entirely with the authors.