arXiv subjects

Jin Xie

Publications and source records attributed to Jin Xie.

112 records · Page 7Linked to original sources

arXiv2018

A Method to Facilitate Cancer Detection and Type Classification from Gene Expression Data using a Deep Autoencoder and Neural Network

With the increased affordability and availability of whole-genome sequencing, large-scale and high-throughput gene expression is widely used to characterize diseases, including cancers. However, establishing specificity in cancer diagnosis using gene expression data continues to pose challenges due to the high dimensionality and complexity of the data. Here we present models of deep learning (DL) and apply them to gene expression data for the diagnosis and categorization of cancer. In this study, we have developed two DL models using messenger ribonucleic acid (mRNA) datasets available from the Genomic Data Commons repository. Our models achieved 98% accuracy in cancer detection, with false negative and false positive rates below 1.7%. In our results, we demonstrated that 18 out of 32 cancer-typing classifications achieved more than 90% accuracy. Due to the limitation of a small sample size (less than 50 observations), certain cancers could not achieve a higher accuracy in typing classification, but still achieved high accuracy for the cancer detection task. To validate our models, we compared them with traditional statistical models. The main advantage of our models over traditional cancer detection is the ability to use data from various cancer types to automatically form features to enhance the detection and diagnosis of a specific cancer type.

stat.ML↗

arXiv2016

Distributions of topological tree metrics between a species tree and a gene tree

In order to conduct a statistical analysis on a given set of phylogenetic gene trees, we often use a distance measure between two trees. In a statistical distance-based method to analyze discordance between gene trees, it is a key to decide "biological meaningful" and "statistically well-distributed" distance between trees. Thus, in this paper, we study the distributions of the three tree distance metrics: the edge difference, the path difference, and the precise $K$ interval cospeciation distance, between two trees: first, we focus on distributions of the three tree distances between two random unrooted trees with $n$ leaves ($n \geq 4$); and then we focus on the distributions the three tree distances between a fixed rooted species tree with $n$ leaves and a random gene tree with $n$ leaves generated under the coalescent process with given the species tree. We show some theoretical results as well as simulation study on these distributions.

q-bio.PE↗

arXiv2015

Stochastic safety radius on Neighbor-Joining method and Balanced Minimal Evolution on small trees

A distance-based method to reconstruct a phylogenetic tree with $n$ leaves takes a distance matrix, $n \times n$ symmetric matrix with $0$s in the diagonal, as its input and reconstructs a tree with $n$ leaves using tools in combinatorics. A safety radius is a radius from a tree metric (a distance matrix realizing a true tree) within which the input distance matrices must all lie in order to satisfy a precise combinatorial condition under which the distance-based method is guaranteed to return a correct tree. A stochastic safety radius is a safety radius under which the distance-based method is guaranteed to return a correct tree within a certain probability. In this paper we investigated stochastic safety radii for the neighbor-joining (NJ) method and balanced minimal evolution (BME) method for $n = 5$.

q-bio.PE↗

arXiv2010

Employing Coded Relay in Multi-hop Wireless Networks

In this paper, we study Coded relay (Crelay) in multi-hop wireless networks. Crelay exploits both partial packets and overhearing capabilities of the wireless nodes, and uses Forward Error Correction code in packet forwarding. When a node overhears a partial packet from an upstream node, it informs the upstream node about the number of parity bytes needed to correct the errors, such that the upstream node need only send a small amount of parity bytes instead of the complete packet, hence improving the network efficiency. Our main contributions include the following. First, we propose an efficient network protocol that can exploit partial packets and overhearing. Second, we study the routing problem in networks with Crelay and propose a greedy algorithm for finding the paths. Third, we propose an error ratio estimator, called AMPS, that can estimate the number of byte errors in a received frame with good accuracy at a low overhead of only 8 bytes per frame, where the estimator is needed for a node to find the number of needed parity bytes. Fourth, we implement the proposed protocol and algorithm within the Click modular router, and our experiments show that Crelay can significantly improve the performance of wireless networks.

cs.NI↗