Search Results

You are looking at 1 - 3 of 3 items for

  • Author: David Lu x
Clear All Modify Search
Open access

Yuan Yuan Lu, Carsten Zorn, David Král, Ming Bai and Xing Ke Yang

Abstract

The small Southeast Asian ruteline genus Glenopopillia Lin, 1980 is revised. We describe four new species: Glenopopillia albopilosa Zorn & Lu sp. nov. from Vietnam, Glenopopillia forceps Zorn & Lu sp. nov. from India, Glenopopillia mengi Lu & Zorn sp. nov. from China and Laos, and Glenopopillia skalei Zorn & Lu sp. nov. from Vietnam; and one new subspecies: Glenopopillia rufipennis nigropicta Zorn & Lu subsp. nov. from Laos; propose two new combinations: Glenopopillia fossulata (Benderitter, 1929) comb. nov. (from Strigoderma fossulata Benderitter, 1929) and Glenopopillia klossi (Ohaus, 1926) comb. nov. (from Spilota klossi Ohaus, 1926), bringing the total number of species group taxa in this genus to ten. We characterize the genus, provide a key to the species, describe and diagnose each species group taxon, and compile a distribution map. A lectotype for Spilota klossi Ohaus, 1926 is designated.

Open access

Ilze Radoviča, Rūdolfs Bērziņs, Gustavs Latkovskis, Dāvids Fridmanis, Liene Ņikitina-Zaķe, Kārlis Ventiņš, Guna Ozola, Andrejs Ērglis and Jānis Kloviņš

Abstract

Familial hypercholesterolemia (FH) is one of the most common single gene disorders, which is mostly inherited as an autosomal dominant trait. The physical signs of FH are elevated low density lipoprotein cholesterol (LDL-C), elevated total cholesterol (TC) levels and tendon xantomas. Identification and early treatment of affected individuals is desirable and in lack of physical symptoms DNA-based diagnosis provides confirmation of diagnosis and enables early patient management. The majority of FH cases are caused by mutations in four genes (APOB, LADLR, PCSK9, and LDLRAP1). There are commercial kits available for testing of the 20 most common FH causing mutations, but the spectrum of disease-causing mutations is quite diverse in various populations and these tests cover only a minority of disease-causing genetic variants. There is therefore a need to determine the full spectrum of mutations in LDLR, APOB, PCSK9, and LDLRAP1 genes in each population. Here we report mutations found in 16 patients with suspected FH in a sample from the Genome Database of the Latvian population enrolled at the Latvian Centre of Cardiology. We used the next generation sequencing approach to determine the full spectrum of mutations in coding regions of LDLR, APOB, PCSK9, and LDLRAP1. In total we found 22 missense mutations, from which only rs5742904 (Arg3527Gln) in APOB gene had been previously described as a FH-causing mutation confirming FH in one patient. Possible FH-causing mutations however, were identified in the majority of patients. The conclusion is that the most commonly employed commercial mutation panel is not sufficient for diagnosis of FH patients and NGS can help to identify FH-causing mutations in the Latvian population.

Open access

Sanjit Bhat, David Lu, Albert Kwon and Srinivas Devadas

Abstract

In recent years, there have been several works that use website fingerprinting techniques to enable a local adversary to determine which website a Tor user visits. While the current state-of-the-art attack, which uses deep learning, outperforms prior art with medium to large amounts of data, it attains marginal to no accuracy improvements when both use small amounts of training data. In this work, we propose Var-CNN, a website fingerprinting attack that leverages deep learning techniques along with novel insights specific to packet sequence classification. In open-world settings with large amounts of data, Var-CNN attains over 1% higher true positive rate (TPR) than state-of-the-art attacks while achieving 4× lower false positive rate (FPR). Var-CNN’s improvements are especially notable in low-data scenarios, where it reduces the FPR of prior art by 3.12% while increasing the TPR by 13%. Overall, insights used to develop Var-CNN can be applied to future deep learning based attacks, and substantially reduce the amount of training data needed to perform a successful website fingerprinting attack. This shortens the time needed for data collection and lowers the likelihood of having data staleness issues.