Bioassay screening of plant extracts can identify unique lead compounds for drug development, but the "hit rate" from random screening is very low. Targeted screening of medicinal plants has been repeatedly reported to increase the percentage of samples displaying bioactivity. Contrarily, Maranz (2012) suggested that African antimalarial plants were unsuitable sources of antimalarial drugs because high prevalence of malaria would result in rapid evolution of resistance to active compounds that directly targeted the parasite. As malaria is highly prevalent in much of Madagascar, it was of interest to determine whether Malagasy antimalarial plants would outperform randomly selected plants in conventional antimalarial assays being conducted as part of a discovery program. Of 1294 plant samples screened for antimalarial activity, 39.6% had an IC50 <50 μg/ml and 21.1% had an IC50 <20 μg/ml (the minimum to qualify as a first-pass "hit"). Ethnobotanical uses were coded at both the generic and the species level, as neither samples nor use reports in literature were always identifiable to species level. The 526 samples belonging to genera having reported uses for malaria were slightly more likely than average to display activity (44.3% with IC50 <50 μg/ml, p < .01; 23.2% with IC50 <20 μg/ml). Of these, 67 samples from individual species with documented use were still more likely to be modestly active (49.3% with IC50 <50 μg/ml), yet less likely to be highly active (17.9% with IC50 <20 μg/ml). Thus, in this specific context, ethnobotanically directed screening would not have substantially improved screening efficiency, and would have missed most of the potential hits.

中文翻译：

---

马达加斯加植物的抗疟用途与抗疟生物测定的性能相关性较差。

植物提取物的生物测定筛选可以识别用于药物开发的独特先导化合物，但随机筛选的“命中率”非常低。药用植物的针对性筛选已被多次报道可以增加显示生物活性的样品的百分比。相反，Maranz（2012）认为非洲抗疟植物不适合作为抗疟药物来源，因为疟疾的高流行会导致对直接针对寄生虫的活性化合物的抗药性迅速进化。由于疟疾在马达加斯加的大部分地区非常流行，因此有兴趣确定马达加斯加的抗疟植物是否会在作为发现计划的一部分进行的传统抗疟测定中优于随机选择的植物。在筛选抗疟活性的 1294 个植物样本中，39.6% 的 IC50 <50 μg/ml，21.1% 的 IC50 <20 μg/ml（符合首次“命中”资格的最低值）。民族植物学用途在属属和物种水平上进行编码，因为文献中的样本和使用报告都不总是可识别到物种水平。属于报告用于疟疾用途的属的 526 个样本显示活性的可能性略高于平均值（44.3%，IC50 <50 μg/ml，p < .01；23.2%，IC50 <20 μg/ml）。其中，来自有使用记录的单个物种的 67 个样品仍然更有可能具有中等活性（49.3%，IC50 <50 μg/ml），但不太可能具有高活性（17.9%，IC50 <20 μg/ml）。因此，在这种特定情况下，民族植物学定向筛选不会显着提高筛选效率，并且会错过大部分潜在的命中。