Personalized Driver Gene Prediction Using Graph Convolutional Networks with Conditional Random Fields.

Cancer is a complex and evolutionary disease mainly driven by the accumulation of genetic variations in genes. Identifying cancer driver genes is important. However, most related studies have focused on the population level. Cancer is a disease with high heterogeneity. Thus, the discovery of driver genes at the individual level is becoming more valuable but is a great challenge. Although there have been some computational methods proposed to tackle this challenge, few can cover all patient samples well, and there is still room for performance improvement. In this study, to identify individual-level driver genes more efficiently, we propose the PDGCN method. PDGCN integrates multiple types of data features, including mutation, expression, methylation, copy number data, and system-level gene features, along with network structural features extracted using Node2vec in order to construct a sample-gene interaction network. Prediction is performed using a graphical convolutional neural network model with a conditional random field layer, which is able to better combine the network structural features with biological attribute features. Experiments on the ACC (Adrenocortical Cancer) and KICH (Kidney Chromophobe) datasets from TCGA (The Cancer Genome Atlas) demonstrated that the method performs better compared to other similar methods. It can identify not only frequently mutated driver genes, but also rare candidate driver genes and novel biomarker genes. The results of the survival and enrichment analyses of these detected genes demonstrate that the method can identify important driver genes at the individual level.

Fine-mapping of qTGW2, a quantitative trait locus for grain weight in rice (Oryza sativa L.).

BACKGROUND: Grain weight is a grain yield component, which is an integrated index of grain length, width and thickness. They are controlled by a large number of quantitative trait loci (QTLs). Besides major QTLs, minor QTLs play an essential role. In our previous studies, QTL analysis for grain length and width was performed using a recombinant inbred line population derived from rice cross TQ/IRBB lines. Two major QTLs were detected, which were located in proximity to GS3 and GW5 that have been cloned. In the present study, QTLs for grain weight and shape were identified using rice populations that were homozygous at GS3 and GW5. METHOD: Nine populations derived from the indica rice cross TQ/IRBB52 were used. An F10:11population named W1, consisting of 250 families and covering 16 segregating regions, was developed from one residual heterozygote (RH) in the F7generation of Teqing/IRBB52. Three near isogenic line (NIL)-F2 populations, ZH1, ZH2 and ZH3 that comprised 205, 239 and 234 plants, respectively, were derived from three RHs in F10:11. They segregated the target QTL region in an isogenic background. Two NIL populations, HY2 and HY3, were respectively produced from homozygous progeny of the ZH2 and ZH3 populations. Three other NIL-F2 populations, Z1, Z2 and Z3, were established using three RHs having smaller heterozygous segments. QTL analysis for 1000-grain weight (TGW), grain length (GL), grain width (GW), and length/width ratio (LWR) was conducted using QTL IciMapping and SAS procedure with GLM model. RESULT: A total of 27 QTLs distributed on 12 chromosomes were identified. One QTL cluster, qTGW2/qGL2/qGW2 located in the terminal region of chromosome 2, were selected for further analysis. Two linked QTLs were separated in region Tw31911-RM266. qGL2 was located in Tw31911-Tw32437 and mainly controlled GL and GW. The effects were larger on GL than on GW and the allelic directions were opposite. qTGW2 was located in Tw35293-RM266 and affected TGW, GL and GW with the same allelic direction. Finally, qTGW2 was delimited within a 103-kb region flanked by Tw35293 and Tw35395. CONCLUSION: qTGW2 with significant effects on TGW, GL and GW was validated and fine-mapped using NIL and NIL-F2 populations. These results provide a basis for map-based cloning of qTGW2 and utilization of qTGW2 in the breeding of high-yielding rice varieties.

Management of type I choledochal cyst in adult: totally laparoscopic resection and Roux-en-Y hepaticoenterostomy.

BACKGROUND/OBJECTIVE: Choledochal cysts are congenital dilations of the biliary tree. The accepted mode of treatment is total excision with hepaticojejunostomy. In this retrospective study, we present our technique and results of laparoscopic choledochal cyst excisions. METHODS: We retrospectively studied 45 patients who had undergone laparoscopic choledochal cyst excision in our institutes from September 2006 to August 2009. Data including age, gender, type of cyst, symptoms, surgical technique, conversion rate, morbidity, and mortality were analyzed. RESULTS: There were type Ic (cystic) choledochal cysts in 31 patients (68.9%) and type If (fusiform) in 14 patients (31.1%). An anomalous pancreaticobiliary duct junction union was found in 66.7%. Forty percent (18 out of 45) and 37.8% (17 out of 45) cases had stones within the cysts and gallbladders, respectively. The average size of the cysts was 40.3 +/- 16.9 cm(2). The mean operative time was 307.7 +/- 58.0 min, the estimated operative blood loss was 252.3 +/- 162.5 ml, and the conversion rate was 8.9%. The mean hospital stay was 8.3 +/- 3.2 days. The overall morbidity rate was 17.1%, the reoperation rate was zero, and the mortality rate was also zero. CONCLUSIONS: Totally, laparoscopic management of type I choledochal cysts, although technically challenging, is safe and feasible in experienced hands.