In vitro and in silico computational methods for assessing vaginal permeability.

OBJECTIVE: Vaginal administration is an important alternative to the oral route for both topical and systemic use. Therefore, the development of reliable in silico methods for the study of drugs permeability is becoming popular in order to avoid time-consuming and costly experiments. METHODS: In the current study, Franz cells and appropriate HPLC or ESI-Q/MS analytical methods were used to experimentally measure the apparent permeability coefficient (Papp) of 108 compounds (drugs and non-drugs). Papp values were then correlate with 75 molecular descriptors (physicochemical, structural, and pharmacokinetic) by developing two Quantitative Structure Permeability Relationship (QSPR) models, a Partial Least Square (PLS) and a Support Vector Machine (SVM). Both were validated by internal, external and cross-validation. RESULTS: Based on the calculated statistical parameters (PLS model A: R2 = 0.673 and Q2 = 0.594, PLS model B: R2 = 0.902 and Q2 = 0.631, SVM: R2 = 0.708 and Q2 = 0.758). SVM presents higher predictability while PLS adequately interprets the theory of permeability. CONCLUSIONS: The most important parameters for vaginal permeability were found to be the relative PSA, logP, logD, water solubility and fraction unbound (FU). Respectively, the combination of both models could be a useful tool for understanding and predicting the vaginal permeability of drug candidates.

CoFly-WeedDB: A UAV image dataset for weed detection and species identification.

The CoFly-WeedDB contains 201 RGB images (∼436 MB) from the attached camera of DJI Phantom Pro 4 from a cotton field in Larissa, Greece during the first stages of plant growth. The 1280 × 720 RGB images were collected while the Unmanned Aerial Vehicle (UAV) was performing a coverage mission over the field's area. During the designed mission, the camera angle was adjusted to -87°, vertically with the field. The flight altitude and speed of the UAV were equal to 5 m and 3 m/s, respectively, aiming to provide a close and clear view of the weed instances. All images have been annotated by expert agronomists using the LabelMe annotation tool, providing the exact boundaries of 3 types of common weeds in this type of crop, namely (i) Johnson grass, (ii) Field bindweed, and (iii) Purslane. The dataset can be used alone and in combination with other datasets to develop AI-based methodologies for automatic weed segmentation and classification purposes.