Biometrical Analysis:

The understanding of ligand - receptor relationships has been a critical step in developing vaccines and therapeutic approaches to immune-mediated infectious and neoplastic diseases. Mixture Sciences has developed a technology that combines the use of Positional Scanning Libraries with a biometrical, score matrix-based prediction strategy in order to quantitatively analyze these relationships. This technology, Biometrical Analysis, will allow the researcher to:

• Identify target receptor specificity.
  
  
• Identify new T-cell ligands for infectious diseases, cancer and autoimmune disorders.

Mixture Sciences' Biometrical Analysis follows the four steps outlined below in order to accomplish the previously mentioned objectives.

Step 1 - Test Positional Scanning Library for biological activity:

The screening data permits the identification of the key functionalities at each diversity position (as shown below). The activity found for a mixture is due to the presence of specific active peptide(s) within the mixture, and not the individual amino acids as separate independent entities. The combination of all positional functional groups identified as key elements leads to active compound(s).

Screening Data: Click to enlarge

Mixture Sciences has an extensive collection of positional scanning libraries ideally designed for use with the Biometrical Analysis.

top

Step 2 - Design matrix based on the activity of each mixture of the positional scanning library:

The screening results are used to generate a scoring matrix that assigns numerical values to the activity of the defined amino acids at each position in the mixtures. In the matrix the columns represent the position of the peptide sequence and the rows represent the 20 amino acids used in that position (see below). The Biometrical Analysis then uses this matrix to score all the decapeptides in a given database. For example the decapeptide stretch Flu HA (308-317) has the sequence YVKQNTLKLA. The Biometrical Analysis then assigns a value for each amino acid in the decapeptide according to its position (Y in P1=30.63, V in P2=15.65, and so forth) for a total score of 256.01.

Scoring Matrix: Click to enlarge

top

Step 3 - Use matrix to score all peptides from proteins contained in database(s):

The Biometrical Analysis then uses the matrix that is generated in Step 2 to score all the decapeptides in a protein database by moving a decamer scoring window across the known protein sequence in one amino acid increments (note: Any length peptide can be used since it is dependent on the positional scanning library. For instance we used a decapeptide positional scanning library to get this data so we will use a decapeptide decamer scoring window. If we used a hexapeptide library, we would use a hexamer window). A predicted stimulatory score is calculated for all the decapeptides. The figure below illustrates the process of scoring three peptides, based on the values of the matrix. The peptides that have the highest score are predicted to have the greatest stimulatory effect on the given receptor.

Overlapping Decapetides: Click to enlarge

After every peptide in the protein database is given a score the researcher can print out a list of all the peptides that were scored. This data contains the rank of each peptide compared to all other peptides scored, the score of each peptide, the sequence of each peptide and the protein the peptide came from. For example in the illustration below the highest scoring peptide sequence identified was -YFKQNSGRLP- with a score of 277.1, this sequence was identified in the Streptococcus thermophilus bacteriophage Sfi21 protein.

In addition to listing the peptides and their scores by rank the data can be represented in the form of a scoring distribution as illustrated below. The below scoring distribution represents more than 23 million peptides scored from one run of the Biometrical Analysis. It can be seen that a relatively small number of peptides have the highest scores. From experiments conducted with the Biometrical Analysis the native ligand has generally been identified as one of the top scoring sequences.

Scoring Distribution: Click to enlarge

top

Step 4 - Synthesize identified peptides and test for activity

The researcher can then select the peptides that will be individually synthesized and test them for activity. Below is a figure that shows the activity of some peptides derived from this analysis. The figure demonstrates that the score generated from the matrix correlates well with the actual activity of the peptide.

Individual Peptides: Click to enlarge

top