Study On The Distribution Patterns And Mechanism Of Nature-derived Drugs Based On The Phylogenetic Tree

Natural products contribute greatly to the modern drug discovery,and more than 60% of approved drugs are natural products or their derivatives/mimics.With the rapid improvemetns of modern science and technology,the generation of a large amount of natural products and related data provides useful resources for natural product-based drug discovery.It is considered that the number of natural products with structure elucidation exceeds 300,000.When it comes to the related research areas,the number is even higher,thus providing many challenges for the quick and efficent selection of natural products or species for further research.In this dissertation,with the predictive power of phylogeny and bioinformatics tools,the distribution patterns of nature-derived drugs in phylogenetic trees were analyzed to seek the species origins of the high “drug-like” natural-product scaffolds.The findings in this study may narrow down the searching scope by focusing only on much smaller amount of compounds or species within “cluster”,and enable more prioritized,focused,rational bioprospecting for novel drug-like natural products.The specific contents as follows.Firstly,50 natural products databases were collected for a comprehensive analysis of their current developmental trends,through the publication year,downloads,website access,and contents.For 35 open source databases,the collection of compounds varied from several hundred to hundreds of thousands,and 82.9% of them provided partial or full data download,and 22.9% of their websites were not stable,and 62.8% of them with species origins.Those findings may guide the construction of better share,public,and comprehensive natural products databases.Then,with the interest in species origin of drug-like natural products,we collected 19261 species(with NCBI taxonmy ID),together with their related 88693 compounds,from those open source databases by data mining tools,including 962 approved drugs(natural products or their derivatives)within 351 species.Moreover,an in-house database with natural products and their species origins was constructed for the following research work.Secondly,260 anti-cancer drugs approved in the past 70 years were comprehensively analyzed.207 out of these 260 drugs derived from or inspired by the natural products were selected for further study.Phylogenetic distribution of those drugs further revealed that 81% of all nature-derived anti-cancer drugs were concentrated in the Bacteria and Viridiplantae kingdoms,within 5 clusters(Actinonycetales,Gammaproteobacteria,commelinids,fabids and lamiids).Moreover,based on their productivity,drug-producing species were categorized into productive(CPS),newly emerging(CNS)and less-productive(CLS).Statistical significances in druglikeness between drugs from CPS and CLS were observed.And there was a clear shift in drug targets from the previous DNA related molecules to the current kinase proteins and their related pathways.Thirdly,given that many modern drugs are derived from plants which were widely used in traditional medicine,traditional knowledge has proven to be a useful tool for novel plant-based drugs.With the predictive power of phylogeny,we conducted a comparative analysis of the distribution patterns between plant-derived drugs and traditional drugs(from China,South Africa,New Zealand,and Nepal)in the phylogenetic tree of Viridiplantae kingdom.Phylogenetic distribution of those drugs revealed 9 “hot” drug-productive overlap clusters,especially multiple species from Ranunculales were reported with high potential to deliver new drugs.And 12 “hot” drug-productive overlap families were further identified,including Asteraceae,Apocynaceae,Convolvulaceae,Cucurbitaceae,Euphorbiaceae,Lamiaceae,Malvaceae,Rosaceae,Rubiaceae,Rutaceae,Sapindaceae,and Solanaceae.Moreover,compared with 24.7% of all PTCM(Plant-based Traditional Chinese Medicine),37.6% of which from the “hot” drug-productive overlap families shared more clinical information at Clinical Trials.gov,demonstrating that “hot” species groups may provide more potential pharmaceutical clues.Finally,based on the above chapters in this thesis,we explored the predictive power of phylogeny for assessing ethnomedical functions of Traditional Chinese medicine(TCM)herbs.252 plant-derived drugs(from 198 species)and 1884 TCMs(from 1676 species)were compared from species,genus and family level,and their overlaps are 98,98,and 64,respectively.99% approved drugs and 70% TCMs were concentrated in 64 families.At species level,from the three aspects of plant species,plant parts,and therapeutic effects,the further comparative study about 68 plant-derived drugs and 93 TCMs,revealed that the corresponding known therapeutic mechanisms of modern drugs may provide useful clues to the mechanisms reseach of 15 ethnomedical(sub)class,which may also promote the TCM-based drug discovery.