Bliss: The Black List Sequence Screening Pipeline

Concerns have been raised that individuals with malicious intent could exploit DNA synthesis technology to obtain genetic elements from pathogenic organisms that would otherwise be difficult to obtain. In response, the U.S. Department of Health and Human Services (HHS) issued the Screening Framework Guidance for Providers of Synthetic Double-Stranded DNA. This voluntary Guidance outlines the U.S. government’s recommendations for screening double-stranded DNA to ensure that existing Select Agent Regulations (SAR) and Export Administration Regulations (EAR) are followed, to encourage best practices in addressing biosecurity concerns, and to reduce the risk that individuals with ill intent could exploit the application of DNA synthesis technology to obtain genetic material derived from or encoding Select Agents or Toxins, or agents on the EAR’s Commerce Control List (CCL).

     In accordance with the HHS guidance, the U.S. Department of Energy Joint Genome Institute (JGI) has developed a DNA screening pipeline (BLiSS – Black List Sequence Screening) to screen all sequences that it synthesizes through its DNA Synthesis Science program. BLiSS detects “sequences of concern” of at least 200 nucleotides in length on either DNA strand, including polypeptide translations using the three alternative reading frames on each DNA strand (six-frame translation).  Sequences to be synthesized are aligned to GenBank’s non-redundant nucleotide and protein databases. A “Best Match” approach is used to determine whether any of the sequences of concern identified within the sequence to be synthesized are unique to Select Agents or Toxins, or CCL-listed agents, toxins or genetic elements, to minimize false positives from closely related organisms or highly conserved “house-keeping genes” which do not pose a biosecurity threat.

     Using the lessons learned by screening over 10,000 sequences submitted to the JGI for synthesis, we are modifying the pipeline to detect when a “sequence of concern” has a high likelihood of being a gene that is not involved in the pathogenicity of the Select Agent. This will greatly aid in the follow-up required when sequences fail the screening process. This follow-up requires human interpretation so it cannot be automated, and has been identified as the most costly aspect of implementing the Guidance by double-stranded DNA providers.