Diagnostic assessment of deep learning for melanocytic lesions using whole-slide pathological images.

BACKGROUND: Deep learning has the potential to improve diagnostic accuracy and efficiency in medical image recognition. In the current study, we developed a deep learning algorithm and assessed its performance in discriminating melanoma from nevus using whole-slide pathological images (WSIs). METHODS: The deep learning algorithm was trained and validated using a set of 781 WSIs (86 melanomas, 695 nevi) from PLA General Hospital. The diagnostic performance of the algorithm was tested on an independent test set of 104 WSIs (29 melanomas, 75 nevi) from Tianjin Chang Zheng Hospital. The same test set was also diagnostically classified by 7 expert dermatopathologists. RESULTS: The deep learning algorithm receiver operating characteristic (ROC) curve achieved a sensitivity 100% at the specificity of 94.7% in the classification of melanoma and nevus on the test set. The area under ROC curve was 0.99. Dermatopathologists achieved a mean sensitivity and specificity of 95.1% (95% confidence interval [CI]: 92.0%-98.2%) and 96.0% (95% CI: 94.2%-97.8%), respectively. At the operating point of sensitivity of 95.1%, the algorithm revealed a comparable specificity with 7 dermatopathologists (97.3% vs. 96.0%, P = 0.11). At the operating point of specificity of 96.0%, the algorithm also achieved a comparable sensitivity with 7 dermatopathologists (96.5% vs. 95.1%, P = 0.30). A more transparent and interpretable diagnosis could be generated by highlighting the regions of interest recognized by the algorithm in WSIs. CONCLUSION: The performance of the deep learning algorithm was on par with that of 7 expert dermatopathologists in interpreting WSIs with melanocytic lesions. By pre-screening the suspicious melanoma regions, it might serve as a supplemental diagnostic tool to improve working efficiency of pathologists.

Stability assessment of a new antithrombotic small peptide, Arg-Gly-Asp-Trp-Arg (RGDWR), and its derivative.

OBJECTIVE: To assess the stabilities of Arg-Gly-Asp-Trp-Arg (RGDWR, designated as RWR), a new patented antithrombotic small peptide, and its derivative with ω-aminocaprylic acid on its N-terminus (ωRWR). RESULTS: RWR in rat plasma was decreased by between 32 and 48% after 4 h incubation on ice, indicating its instability in plasma. In contrast, ωRWR in plasma remained at 96-107%. Concentration changes were within 6.2% for ωRWR after storage in various conditions. ωRWR is therefore stable in rat plasma, as well as under different storage methods. Furthermore, ω-aminocaprylic acid added onto the RWR peptide did not affect its antiplatelet aggregation activity. CONCLUSIONS: A novel small peptide, ωRWR, has been developed with a good stability for possible antithrombotic use.

Development of novel, 384-well high-throughput assay panels for human drug transporters: drug interaction and safety assessment in support of discovery research.

Transporter proteins are known to play a critical role in affecting the overall absorption, distribution, metabolism, and excretion characteristics of drug candidates. In addition to efflux transporters (P-gp, BCRP, MRP2, etc.) that limit absorption, there has been a renewed interest in influx transporters at the renal (OATs, OCTs) and hepatic (OATPs, BSEP, NTCP, etc.) organ level that can cause significant clinical drug-drug interactions (DDIs). Several of these transporters are also critical for hepatobiliary disposition of bilirubin and bile acid/salts, and their inhibition is directly implicated in hepatic toxicities. Regulatory agencies took action to address transporter-mediated DDI with the goal of ensuring drug safety in the clinic and on the market. To meet regulatory requirements, advanced bioassay technology and automation solutions were implemented for high-throughput transporter screening to provide structure-activity relationship within lead optimization. To enhance capacity, several functional assay formats were miniaturized to 384-well throughput including novel fluorescence-based uptake and efflux inhibition assays using high-content image analysis as well as cell-based radioactive uptake and vesicle-based efflux inhibition assays. This high-throughput capability enabled a paradigm shift from studying transporter-related issues in the development space to identifying and dialing out these concerns early on in discovery for enhanced mechanism-based efficacy while circumventing DDIs and transporter toxicities.

Case study: technology initiative led to advanced lead optimization screening processes at Bristol-Myers Squibb, 2004-2009.

In this paper, we review the key solutions that enabled evolution of the lead optimization screening support process at Bristol-Myers Squibb (BMS) between 2004 and 2009. During this time, technology infrastructure investment and scientific expertise integration laid the foundations to build and tailor lead optimization screening support models across all therapeutic groups at BMS. Together, harnessing advanced screening technology platforms and expanding panel screening strategy led to a paradigm shift at BMS in supporting lead optimization screening capability. Parallel SAR and structure liability relationship (SLR) screening approaches were first and broadly introduced to empower more-rapid and -informed decisions about chemical synthesis strategy and to broaden options for identifying high-quality drug candidates during lead optimization.