Professor Li Guojun's team republishes important result of transcriptome assembly

Recently, Professor Li Guojun's team published the latest results about the transcriptome reconstruction "TransLiG: ade novotranscriptome assembler that uses line graph iteration" in the top international journal Genome Biology (IF: 13.481). The dissertation was independently completed by Shandong University, with Liu Juntao, a young teacher from the School of Mathematics and Statistics of Weihai Campus, and Yu Ting, a doctoral student from the Central Campus, as co-authors, and Professor Li Guojun as the corresponding author.

RNA-seq sequencing technology is a powerful tool for collecting gene expression data at the full transcriptome level, with unprecedented sensitivity and accuracy. Compared to microarray chips and EST sequencing, RNA-seq achieves single-nucleotide resolution, has a higher dynamic range, and allows reliable identification of rare transcripts and alternative splicing. Therefore, this technology provides important technical support for the treatment of human complex diseases and the study of the function of new species. However, how to use this vast amount of RNA-seq data to calculate and reconstruct the transcriptome has become a huge challenge around the world. At present, this problem has not been well modeled, thus becoming a key bottleneck in the analysis of RNA-seq data in this field.

To this end, Li Guojun's team has developed a new transcriptome de novo assembly algorithm, TransLiG. The algorithm first discovered a new technique for repairing the current severe fragmentation of gene splicing graphs, enabling a large number of low-coverage transcripts to be accurately spliced out, which is not possible achieved by current assembly algorithms. Based on the repaired splicing graph, a new phasing and combing technology is designed, and the sequence depth information and the double-end sequencing information are organically fused in the entire assembly process using the iteration of line graph, the theoretical global optimization is obtained, so that its computational prediction performance achieves a qualitative leap. Test evaluation shows that TransLiG clearly surpasses the most popular assembly algorithm. For example, compared to Trinity (Nature Biotechnology, 2011), the most commonly used tool, in the RNA-seq test data of a group of mice, TransLiG has improved the sensitivity by 15.6% and at the same time, the accuracy has increased by nearly 15%. Not only that, TransLiG can maintain stable and optimal performance under different evaluation parameters.

This research work is an important breakthrough in the field of transcriptome splicing and will undoubtedly play an important role in promoting subsequent research in related fields. Professor Li Guojun and Dr. Liu Juntao have published 3 academic papers in Genome Biology. This research was supported by a key project of the National Natural Science Foundation of China.

(Editor in Charge: Lei Hao)