Research Interest



The novel view of cancer

Previous opinions on cancer considered cancer as driven by specific somatic mutations. However, cancer studies revealed that most cancer are not caused by specific driver mutations, but caused by complex gene networks.

We have found that the translation is globally regulated in cancer. This may serve as a signature of most cancer types. Cancer cells are able to regulate the translatome of their microenvironment and thus gain the survival and progression. Mutations may be simply a result of translatome misregulation rather than the reason. This may raise a revolution on cancer studies and therapy.

Fight against the antibiotic-resistant bacteria

Antibiotic-resistant (AR) bacteria is becoming a major threat on public health, which causes 85% deaths of infectious diseases. It is increasingly difficult to develop new antibiotics. In contrast, the bacteria evolves resistance within one year of the clinical usage. Therefore, we try to answer the question: why can the “simple” bacteria evolve resistance against everything?

We discovered that translation regulation is a swift and common defense mechanism against most antibiotics types. Manipulation of the translation system can further tune this response. Therefore, translation system may serve as a promising target to prevent the resistance.

We are also willing to collaborate with the structural biologists and pharmaceutical scientists to design specific antibiotics against prokaryotic ribosomes to fight against the AR bacteria.

The Human Proteome Project

Translatome has been adaopted as one of the key resource pillar of the Human Proteome Project (HPP). As an important group in the HPP, we are still developing new methods, both experimental and computational, aiming the complete resolution of the human proteome, including the protein types (especially the protein-coding “non-coding RNAs”), sequence variants, modifications and dynamics. Most of our current ongoing work showed potential of groundbreaking advancements. Almost all of such improvements are based on the translatome-aided proteomics strategies.

Finding biomarkers for complex diseases

Taking the advantage of our omics infrustructure and hyper-accurate algorithms, we are trying to find novel biomarkers, mainly the proteins and exosome-encapsulated nucleic acids, for the diagnosis, prognosis, monitor, etc. of complex diseases, such as cancer, cardiovascular diseases and autoimmune diseases. Using novel statistical design and accurate omics methods, we have successful experiences on the biomarkers of Kawasaki Diseases and promising results of various types of cancer. We are also cutting the cost to a very low level so that these tests may be widely and regularly performed as population-level screening.

Rational design of translational pausing for biochemical engineering

The production of many pharmaceutical and industrial proteins in prokaryotic hosts is hindered by the insolubility of industrial expression products resulting from misfolding. Even with a correct primary sequence, an improper translation elongation rate in a heterologous expression system is an important cause of misfolding. The actual reason is that the correct amino acid sequence do not guarantee the correct folding, even in the same environment. The translational pausing, mediated by the codon selection and tRNA abundance, is an important determinant for protein folding.

Therefore, we established a novel strategy to efficiently promote the expression of soluble and active proteins without altering the amino acid sequence or expression conditions. This strategy uses the rational design of translational pausing based on structural information solely through synonymous substitutions, i.e. no change on the amino acids sequence. We demonstrated this strategy on several proteins which are hard to expressed in soluble form in wild-type sequences. We enhanced the soluble expression yield hundreds to 2000 folds, and reached even higher specific activity than the wild-type. Therefore, this strategy enables new possibilities for the efficient production of bioactive recombinant proteins.

Watch the TED talk by the inventor of FANSe, Dr. Gong Zhang, to recognize how the accurate FANSe outperformed the other algorithms and really saved lives.








We are the only team in the world who developed high-accuracy algorithm solutions for both next-generation sequencing and shotgun mass-spectrometry-based proteomics. We have established the world’s most powerful commercialized cloud-based omics computational platform, the “Chi-Cloud” operated by Chi-Biotech Co. Ltd., for the new era of omics research and the precision health/medicine.

For next-generation sequencing: FANSe series mapping algorithm


  • FANSe 1 (Nucleic Acids Res. (2012) 40(11):e83)
  • FANSe 2 (PLoS One (2014), 9(4):e94250)
  • FANSe 2 splice (Scientific Reports (2017) 7:1053)
  • FANSe 3 (Nucleic Acids Res. (2018) 46(D1), D206)
  • Striking features

  • Extreme speed: whole genome in 1 hour
  • Guaranteed accuracy: mathematical evaluated mismapping rate < 1e-6
  • High experimental verifiability: saves your time for validations and gene pickups for deepened studies.
  • High error tolerance: perfectly supports up to 12% deviation from the ref sequence.
  • Designed for cloud: fully optimized for cloud-based server infrustructure to maximize parallelizability.
  • For shotgun proteomics: Comprehensive algorithms & strategies

    MS identification algorithms

  • ProVerB (J Proteome Res. (2013) 12 (1), 328–335)
  • Dispec (PLoS One (2013) 8 (5), e62724)
  • Strategies for higher sensitivity and confidence

  • Iterative genome correction (J Proteome Res. (2014), 13 (6), 2724–34)
  • Protein-level integration (J Proteome Res. (2017) 16(12), 4446–4454)
  • Translatome-aided proteomics

  • New proteins encoded by non-coding RNAs (Science China Life Sciences (2014) 57 (3), 358-360)
  • Key resource pillar of Human Proteome Project (J Proteome Res. (2014), 13 (1), 50-59)
  • For cloud-based omics data processing: world's most powerful Chi-Cloud

    Striking features

    Extreme capacity
    backed-up by the commercialized private cloud and the Galaxy-II Supercomputer (former No.1 supercomputer in the world), the Chi-Cloud may analyze the WGS data of the entire world population within 1 year.
    Swift data transfer
    patented compression algorithm to allow transfer RNA-seq datasets within minutes via internet.
    One-click analysis
    mapping, quantification, SNV/CNV analysis, differential gene expression, pathway enrichment, ... all in one click.
    Friendly reports
    HTML interactive reports and high-resolution figures. Medical version may directly provide medical indications.