Because the prognosis of melanoma is challenging and inaccurate when using current clinical approaches, clinicians are seeking more accurate molecular markers to improve risk models. Accordingly, we performed a survival analysis on 404 samples from The Cancer Genome Atlas (TCGA) cohort of skin cutaneous melanoma. Using our recently developed gene network model, we identified biological signatures that confidently predict the prognosis of melanoma (p-value < 10−5). Our model predicted 38 cases as low–risk and 54 cases as high–risk. The probability of surviving at least 5 years was 64% for low–risk and 14% for high–risk cases. In particular, we found that the overexpression of specific genes in the mitotic cell cycle pathway and the underexpression of specific genes in the interferon pathway are both associated with poor prognosis. We show that our predictive model assesses the risk more accurately than the traditional Clark staging method. Therefore, our model can help clinicians design treatment strategies more effectively. Furthermore, our findings shed light on the biology of melanoma and its prognosis. This is the first in vivo study that demonstrates the association between the interferon pathway and the prognosis of melanoma.

No abstract

K-Nearest Neighbor (kNN)-based deep learning methods have been applied to many applications due to their simplicity and geometric interpretability. However, the robustness of kNN-based deep classification models has not been thoroughly explored and kNN attack strategies are underdeveloped. In this paper, we first propose an Adversarial Soft kNN (ASK) loss for developing more effective kNNbased deep neural network attack strategies and designing better defense methods against them. Our ASK loss provides a differentiable surrogate of the expected kNN classification error. It is also interpretable as it preserves the mutual information between the perturbed input and the in-class-reference data. We use the ASK loss to design a novel attack method called the ASK-Attack (ASK-Atk), which shows superior attack efficiency and accuracy degradation relative to previous kNN attacks on hidden layers. We then derive an ASK-Defense (ASK-Def) method that optimizes the worst-case ASK training loss. Experiments on CIFAR-10 (ImageNet) show that (i) ASK-Atk achieves ≥ 13% (≥ 13%) improvement in attack success rate over previous kNN attacks, and (ii) ASK-Def outperforms the conventional adversarial training method by ≥ 6.9% (≥ 3.5%) in terms of robustness improvement. Relevant codes are available at https://github.com/wangren09/ASK.