Biology subjects

Gao, G.

Publications and source records attributed to Gao, G..

4 recordsLinked to original sources

bioRxiv2018

Gp78 E3 ubiquitin ligase mediates both basal and damage-induced mitophagy

Mitophagy, the elimination of mitochondria by the autophagy machinery, evolved to monitor mitochondrial health and maintain mitochondrial integrity. PINK1 is a sensor of mitochondrial health that recruits Parkin and other mitophagy-inducing ubiquitin ligases to depolarized mitochondria. However, mechanisms underlying mitophagic control of mitochondrial homeostasis, basal mitophagy, remain poorly understood. The Gp78 E3 ubiquitin ligase, an endoplasmic reticulum membrane protein, induces mitochondrial fission, endoplasmic reticulum-mitochondria contacts and mitophagy of depolarized mitochondria. CRISPR/Cas9 knockout of Gp78 in HT-1080 fibrosarcoma cells results in reduced ER-mitochondria contacts, increased mitochondrial volume and resistance to CCCP-induced mitophagy. Knockdown (KD) of the essential autophagy protein ATG5 increased mitochondrial volume of wild-type cells but did not impact mitochondrial volume of Gp78 knockout cells. This suggests that endogenous Gp78 actively eliminates mitochondria by autophagy in wild-type HT-1080 cells. Damage-induced mitophagy of depolarized mitochondria, in the presence of CCCP, but not basal mitophagy was prevented by knockdown of PINK1. This suggests that endogenous Gp78 plays dual roles in mitophagy induction: 1) control of mitochondrial homeostasis through mitophagy of undamaged mitochondria; and 2) elimination of damaged mitochondria through PINK1.

cell biology↗

bioRxiv2018

Genome-wide association analysis with a 50K transcribed gene SNP-chip identifies QTL affecting muscle yield in rainbow trout

Detection of coding/functional SNPs that change the biological function of a gene may lead to identification of putative causative alleles within QTL regions and discovery of genetic markers with large effects on phenotypes. Two bioinformatics pipelines, GATK and SAMtools, were used to identify ~21K transcribed SNPs with allelic imbalances associated with important aquaculture production traits including body weight, muscle yield, muscle fat content, shear force, and whiteness in addition to resistance/susceptibility to bacterial cold-water disease (BCWD). SNPs were identified from pooled RNA-Seq data collected from ~620 fish, representing 98 families from growth- and 54 families from BCWD-selected lines with divergent phenotypes. In addition, ~29K transcribed SNPs without allelic-imbalances were strategically added to build a 50K Affymetrix SNP-chip. SNPs selected included two SNPs per gene from 14K genes and ~5K non-synonymous SNPs. The SNP-chip was used to genotype 1728 fish. The average SNP calling-rate for samples passing quality control (QC; 1,641 fish) was [≥] 98.5%. Genome-wide association (GWA) study on 878 fish (representing 197 families from 2 consecutive generations) with muscle yield phenotypes and genotyped for 35K polymorphic markers (passing QC) identified several QTL regions explaining together up to 28.40% of the additive genetic variance for muscle yield in this rainbow trout population. The most significant QTLs were on chromosomes 14 and 16 with 12.71% and 10.49% of the genetic variance, respectively. Many of the annotated genes in the QTL regions were previously reported as important regulators of muscle development and cell signaling. No major QTLs were identified in a previous GWA study using a 57K genomic SNP chip on the same fish population. These results indicate improved detection power of the transcribed gene SNP-chip in the target trait and population, allowing identification of large-effect QTLs for important traits in rainbow trout.

genomics↗

bioRxiv2017

Genome-wide Association Studies Reveal Similar Genetic Architecture with Shared and Unique QTL for Bacterial Cold Water Disease Resistance in Two Rainbow Trout Breeding Populations

Bacterial cold water disease (BCWD) causes significant mortality and economic losses in salmonid aquaculture. In previous studies, we identified moderate-large effect QTL for BCWD resistance in rainbow trout (Oncorhynchus mykiss). However, the recent availability of a 57K SNP array and a genome physical map have enabled us to conduct genome-wide association studies (GWAS) that overcome several experimental limitations from our previous work. In the current study, we conducted GWAS for BCWD resistance in two rainbow trout breeding populations using two genotyping platforms, the 57K Affymetrix SNP array and restriction-associated DNA (RAD) sequencing. Overall, we identified 14 moderate-large effect QTL that explained up to 60.8% of the genetic variance in one of the two populations and 27.7% in the other. Four of these QTL were found in both populations explaining a substantial proportion of the variance, although major differences were also detected between the two populations. Our results confirm that BCWD resistance is controlled by the oligogenic inheritance of few moderate-large effect loci and a large-unknown number of loci each having a small effect on BCWD resistance. We detected differences in QTL number and genome location between two GWAS models (weighted single-step GBLUP and Bayes B), which highlights the utility of using different models to uncover QTL. The RAD-SNPs detected a greater number of QTL than the 57K SNP array in one population, suggesting that the RAD-SNPs may uncover polymorphisms that are more unique and informative for the specific population in which they were discovered.

genetics↗

bioRxiv2017

An exact transformation of convolutional kernels applied directly to DNA/RNA sequences

MotivationConvolutional neural network (CNN) has been widely used in functional motifs identification for large-scale DNA/RNA sequences. Currently, however, the only way to interpret such a convolutional kernel is a heuristic construction of a position weight matrix (PWM) from fragments scored highly by that kernel.\n\nResultsInstead of using heuristics, we developed a novel, exact kernel-to-PWM transformation whose equivalency is theoretically proven: the log-likelihood of the resulting PWM generating any DNA/RNA sequence is exactly the sum of a constant and the convolution of the original kernel on the same sequence. Importantly, we further proved that the resulting PWMs performance on sequence classification/regression can be exactly the same as the original kernels under popular CNN frame-works. In simulation, the exact transformation rivals or outperforms the heuristic PWMs in terms of classifying sequences with sequence- or structure-motifs. The exact transformation also faithfully reproduces the output of CNN models on real-world cases, while the heuristic one fails, especially on the case with little prior knowledge on the form of underlying true motifs. Of note, the time complexity of the novel exact transformation is independent on the number of input sequences, enabling it to scale well for massive training sequences.\n\nAvailabilityPython scripts for the transformation from kernel to PWM, the inverted transformation from PWM to kernel, and a proof-of-concept for the maximum likelihood estimation of optimal PWM are available through https://github.com/gao-lab/kernel-to-PWM.\n\nContactgaog@mail.cbi.pku.edu.cn

bioinformatics↗