Microbiome Startup Interface List

Determine the Lab that will processing the data and contract

There are many aspects that needs to be clarified:

  • What certification does the lab has? Often labs are certified for “research purposes” and cannot be used in a clinical setting. This can be critical when you are trying to sell your product to medical professionals (medical insurance and government regulatory groups may require a specific certification).
  • Is the report based on Shotgun or 16s or other measuring technologies?
  • With their report give NCBI numbers or some other identifier.
    • If other identifiers, you will need to convert them to NCBI numbers (a messy area)
  • Will the lab report the percentage of each bacteria (usually this is an easy yes)
  • Will the lab report the percentile of each bacteria across those that report that bacteria in the reference set. A reference set of 1000 samples from healthy individuals is recommended.
  • Will the delivery of data to you by HIPAA complaint?
  • Turnaround time from sample receipt until data is available. Remember, your clients are likely impatient.

Determine the Modifiers to Consider

The response back will be for all modifiers (by number). A list of modifier with names is provided (and can be dynamically obtained from the site to identifier new ones). If is suggested that you assign names familiar to your customers

Determine the type of Inferences you want to use

See Taxonomy Inference with the Microbiome. You can allow all we are currently using OR be “nit picking” for example, for family taxonomy rank use only child inferences. Inference is used to generate reasonable data when we lack any studies.

Determine the Algorithm

MP has its preferred proprietary algorithms which will be used by default. If you have a preferred way of computing weight, we will gladly implement it. This gives the option of claiming your own proprietary algorithms for marketing or for investors.

Show Evidence or Not?

In general, the evidence used to make suggestions are using studies on the US National Library of Medicine Library, These are identified by a PMID number. In a few rare occasion, other sources are used and a DOI identifier will be used. Telling the difference is easy — one is an integer (PMID) and one is not.

The evidence trail will be shown as n-tuple consisting of:

  • ncbi number (bacteria identifier)
  • modifier Id
  • Increase or Decrease ( a + or a – )
  • Source:
    • Integer: PMID
    • Not Integer DOI

Example:

  • (Bacillus, Clostridium butyricum Probiotic, ‘+’, 31815958) : Link

or

Display Ownership that is the Client’s

  1. The name of the bacteria shown on their site (names are often changing)
  2. The name of the modifier shown on their site (often different names for different audiences)
  3. The display of the study (evidence)
    • NCBI provides several public APIs that allow programmatic access to many databases and tools.

Bottom Line

The above are the fundamentals. With symptom capture and sufficient data, we can assist in providing guidance for selecting key bacteria for various symptoms and the ranges that are critical.