Summary In many supervised learning applications, understanding and visualizing the effects of the predictor variables on the predicted response is of paramount importance. A shortcoming of black box supervised learning models (e.g. complex trees, neural networks, boosted trees, random forests, nearest neighbours, local kernel‐weighted methods and support vector regression) in this regard is their lack of interpretability or transparency. Partial dependence plots, which are the most popular approach for visualizing the effects of the predictors with black box supervised learning models, can produce erroneous results if the predictors are strongly correlated, because they require extrapolation of the response at predictor values that are far outside the multivariate envelope of the training data. As an alternative to partial dependence plots, we present a new visualization approach that we term accumulated local effects plots, which do not require this unreliable extrapolation with correlated predictors. Moreover, accumulated local effects plots are far less computationally expensive than partial dependence plots. We also provide an R package ALEPlot as supplementary material to implement our proposed method.

No abstract

BackgroundWe previously reported a PI3K inhibitor S14161 which displays a promising preclinical activity against multiple myeloma (MM) and leukemia, but the chiral structure and poor solubility prevent its further application.MethodsSix S14161 analogs were designed based on the structure–activity relationship; activity of the compounds in terms of cell death and inhibition of PI3K were analyzed by flow cytometry and Western blotting, respectively; anti-myeloma activity in vivo was performed on two independent xenograft models.ResultsAmong the six analogs, BENC-511 was one of the most potent compounds which significantly inhibited PI3K activity and induced MM cell apoptosis. BENC-511 was able to inactivate PI3K and its downstream signals AKT, mTOR, p70S6K, and 4E-BP1 at 1 μM but had no effects on their total protein expression. Consistent with its effects on PI3K activity, BENC-511 induced MM cell apoptosis which was evidenced by the cleavage of Caspase-3 and PARP. Notably, addition of insulin-like growth factor 1 and interleukin-6, two important triggers for PI3K activation in MM cells, partly blocked BENC-511-induced MM cell death, which further demonstrated that PI3K signaling pathway was critical for the anti-myeloma activity of BENC-511. Moreover, BENC-511 also showed potent oral activity against myeloma in vivo. Oral administration of BENC-511 decreased tumor growth up to 80% within 3 weeks in two independent MM xenograft models at a dose of 50 mg/kg body weight, but presented minimal toxicity. Suppression of BENC-511 on MM tumor growth was associated with decreased PI3K/AKT activity and increased cell apoptosis.ConclusionsBecause of its potent anti-MM activity, low toxicity (LD50 oral >1.5 g/kg), and easy synthesis, BENC-511 could be developed as a promising agent for the treatment of MM via suppressing the PI3K/AKT signaling pathway.

BackgroundWe and others have previously shown that the STAT3 signaling pathway is activated in some esophageal squamous cell carcinoma (ESCC) cells and is required for the survival and growth of these primary ESCC-derived xenografts. It has also been shown that the natural polyphenol curcumin is an effective anti-tumor agent.MethodsLuciferase assay and immunoblotting were performed to examine whether curcumin suppressed STAT3 signaling. CCK-8 assay and xenografts were utilized for analyzing ESCC cell growth in culture and mice. Soft agar assay was carried out to determine the colony formation ability of ESCC cells in the presence or absence of curcumin. Cell death and cell cycle were assessed by In CELL Analyzer 2000. Immunohistochemistry and TUNEL assay were used for detecting apoptosis in ESCC tisuses. Molecular docking was performed to evaluate the interaction of curcumin with JAK2. JAK2 activity was assessed using an in vitro cell-free system. HE staining was used to evaluate the ESCC tissues.ResultsThe natural polyphenol curcumin inhibited STAT3 phosphorylation rapidly and blocked STAT3-mediated signaling in ESCC cells. It also induced growth arrest and apoptosis in cultured ESCC cells, which were attenuated by enforced expression of STAT3. Furthermore, curcumin preferentially blocked the growth of primary ESCC-derived xenografts that harbored activated STAT3.ConclusionsCurcumin is able to exert anti-tumor action through inhibiting the STAT3 signaling pathway. Giving its wide use in traditional medicines with low toxicity and few adverse reactions, it is conceivable that curcumin might be further explored as a unique STAT3 inhibitor for anti-cancer therapies.

Background: Clioquinol is a potent chelator of divalent metal ions including zinc. Results: Clioquinol fits the zinc-centered active pocket of HDACs and inhibits HDAC activity, which results in cell cycle arrest and cancer cell apoptosis. Conclusion: Clioquinol inhibits HDAC activity and induces blood cancer cell death. Significance: This is the first report to demonstrate that clioquinol inhibits HDAC activity.