Malaria is a potentially life-threatening parasitic disease, causing serious illness and death worldwide. According to recent data, an estimated 249 million cases of malaria were reported in 2022 across 85 malaria-endemic countries and regions. This figure represents five million increase in cases compared to the previous year 2021 (World Health Organization, 2023). The disease is transmitted to humans through infected female Anopheles mosquitoes, and the main cause of the disease is a protozoan parasite of the genus Plasmodium and phylum Apicomplexa. Among the five Plasmodium species known to infect humans, Plasmodium falciparum accounts for over 90 % of malaria-related deaths (Zekar and Sharman, 2023). Artemisinin-based therapies reduced global malaria fatalities by nearly half between 2000 and 2015. However, mortality rates have risen in recent years due to emerging resistance to antimalarial drugs. From 2010 to 2022, several studies were undertaken by the WHO in different malaria-endemic regions. They reported a higher level of treatment failure of the artemisinin-based combination therapy (ACT) against P. falciparum (World Health Organization, 2023). Therefore, developing next-generation antimicrobial medications with distinct mechanisms of action is crucial.

There are 82 functional eukaryotic PKs reported in P. falciparum with the help of large-scale proteomics and microarray studies. During the asexual blood stages of Plasmodium, PKs phosphorylate both host proteins and endogenous parasitic substrates (Anamika et al., 2005). Studies have demonstrated protein kinase inhibitors significantly inhibit both invasion and intraerythrocytic-stage development of P. falciparum (Dluzewski and Garcia, 1996). Protein kinases are versatile and involved in many stages of the parasitic life cycle; they make good targets for therapeutics. Again, the apicomplexan kinome consists of many groups of eukaryotic kinase enzymes that are not found in their human counterparts due to a large phylogenetic distance. This suggests that selectively inhibiting these enzymes could potentially eliminate the parasite (Cassiano et al., 2021). Kinase targets are identified using two primary techniques. The first technique generates in vitro resistance following whole genome sequencing (McNamara et al., 2013), while the second utilizes chemoproteomics with kinobead technology (Eberl et al., 2019). This method involves immobilizing kinase inhibitors onto sepharose beads to enable pull-down of kinase targets from Plasmodium lysate. Key targets identified through these approaches include PfPI4Kβ, PfPKG, PfGSK3, PfCDPK, and PfPK6 (Vanaerschot et al., 2020; Kato et al., 2008).

Among the various screened functional targets, there has been relatively limited research conducted on PfPK6 (PF3D7_1337100), a serine/threonine protein kinase, classified within the CMGC group due to its shared characteristics with both CDKs and MAPKs (Ward et al., 2004). Unlike typical cyclin-dependent kinases, PfPK6 displays kinase activity independently of cyclin. PfPK6 plays a crucial role in the regulation of the intraerythrocytic cell cycle, promoting the growth and replication of parasites within red blood cells (RBCs). Research has shown that it phosphorylates MAL7P1.38, a regulator of chromosomal condensation, as well as PF10_0047, an RNA-binding protein (Cummins, 2016). This suggests that PfPK6 may play a role in the transcription and translation of genes critical for the parasite's survival and proliferation. Additionally, PfPK6 interacts with other proteins, enabling the parasite to adapt to its environment and effectively evade the host immune response.

CDK2 (PDB Id: 1HCK), the human homologous of PfPK6 contain a smaller amino-terminal lobe that is composed of beta-sheets and the PASTAIRE helix. Some of the amino acid residues involved in cyclin binding activity present in these regions are absent in PfPK6; however, it retains essential regulatory motifs and phosphorylation sites (Thr 14, Tyr 15, Thr 160 in CDKs). The PASTAIRE motif is replaced by SKCILRE sequence in P. falciparum PK6. PfPK6 must utilize an alternative mechanism for cyclin-independent activation to prevent substrate obstruction by the T loop at the catalytic pocket entrance. The protein kinase 6 contains 15 conserved amino acid residues, including the glycine loop in the ATP binding subdomain and lysine in the catalytic subdomain, of eukaryotic kinases. The protein is present in both the cytoplasm and the nucleus, and it undergoes shuttling between these two cellular compartments throughout different stages of the parasite life cycle to execute distinct functions (Bracchi-Ricard et al., 2000). PfPK6 expression occurs during the trophozoite and early schizont stages in the asexual blood stage of P. falciparum, from late G1 to mitosis. Protein kinases are responsible for transferring the gamma phosphate groups from ATP to specific amino acid residues within a protein substrate's side chain. This process plays a crucial role in regulating various biological processes by influencing the structure and function of the proteins involved. The active site of PfPK6 encompasses multiple conserved motifs, such as the glycine-rich loop, catalytic loop, and activation loop, facilitating ATP and substrate binding, as well as phosphate transfer, respectively (Waters and Geyer, 2003).

A recent study verified that 'compound 79′, Ki8751 is a newly identified inhibitor for PfPK6 with an IC50 of less than 5 nM. This compound demonstrates anti-plasmodial activity against P. falciparum in the asexual blood stage, with an EC50 of 39 nM (Ong et al., 2023; Crowther et al., 2016) performed the screening of nearly 14000 small molecules for pharmacological inhibition of five kinases, including PfPK6, and identified eight compounds with potential inhibition up to 138 nM. Other study have reported IKK16 as a dual inhibitor of PfPK6 and PfGSK3. A derivative of IKK16, referred to as ‘compound-23d’, shows potential inhibition of both kinases with IC50 values of 11 nM for PfPK6 and 172 nM for PfGSK3. Furthermore, it exhibits antimalarial activity against the asexual blood stage with an EC50 of 552 nM (Galal et al., 2022). Despite promising in vitro results, these inhibitors have not demonstrated sufficient clinical efficacy in treating malaria in vivo. The emergence of resistance to these inhibitors is a significant concern, similar to what has been observed with other antimalarial drugs. It is crucial to consider pharmacokinetic properties to validate the effectiveness of these inhibitors in vivo.

This study aims to identify a more potent inhibitor with increased selectivity towards PfPK6 using a virtual screening computational approach. For the first time, we have utilized diversified datasets of antimalarials alongside a structure-based approach and a deep learning model to enhance the accuracy of identifying more specific inhibitors. The structure-based approach involves the docking of three-dimensional structures of the target protein and ligand to analyze the binding affinity of the resultant complexes (Vasudevan et al., 2023). In contrast, DL-based screening employs neural networks arranged in multiple layers to predict the interactions between drugs and their targets. This approach requires training on well-defined experimental data to ensure accuracy and reliability (Sharma et al., 2024).

This in-silico methodology facilitates the rapid assessment of millions of drug candidates by accurately predicting their binding affinities to a specific target protein (Verma et al., 2024). By avoiding the necessity of laboratory synthesis and testing of extensive datasets, this approach significantly diminishes both the cost and time associated with the drug discovery pipeline. Furthermore, it provides a streamlined mechanism to evaluate existing FDA-approved drugs against new potential targets (Verma et al., 2024a, Verma et al., 2024b).

Recent advancements in the integration of machine learning (ML) and deep learning (DL) algorithms into drug screening processes have markedly improved prediction accuracy. Moreover, virtual screening extends beyond the examination of binding interactions and affinities, as it is imperative to investigate the pharmacokinetic properties of candidates through ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) analysis (Verma et al., 2021). Key parameters such as bioactivity, intestinal absorption, drug distribution, metabolism, excretion, toxicity, and potential drug-drug interactions must be computed to comprehensively evaluate and screen the compounds under consideration.

Due to its essential role in the parasite's life cycle, unique cyclin-independent activity, structural insights into its binding sites, and interactions with inhibitors, PfPK6 is a promising therapeutic target. This computational approach seeks to develop a novel PfPK6 inhibitor using various in silico techniques and screening methods.