Our Frequently Asked Questions

The Aptamarker platform is a different approach to NGS-based proteomics. We do not rely on pretrained probes on recombinant proteins. Instead, we characterize proteins as they exist in vivo, enabling the characterization of changes to proteins based on post-translational modifications, and multimeric structures.

This is enabled by the following innovations:

FRELEX: translation of protein abundance to DNA sequences using competitor antisense sequences. We use antisense sequences immobilized on resin to act as competitors for Aptamarker binding. The more abundant a target is, the less Aptamarker is captured by the antisense. A microcentrifugation step removes the Aptamarker hybridized to the resin, and the supernatant carrying each Aptamarker bound to a protein is used for NGS analysis

Neomer library: We have reinvented aptamer development by reducing the number of random nucleotides while increasing the structural diversity of the library. We currently apply a library of the same 268 million Aptamarkers to each sample.

Structure analysis: 268M sequences are too many for robust NGS analysis. We partnered with NVIDIA for GPU support and have predicted the structures of all 268M sequences at various temperatures and salt conditions. At room temperature we have 428,686 unique structures, which means that on average 600 different Aptamarker sequences are creating identical structures. We have built code which assigns NGS reads to structure reads, and the analysis is performed at this level.

Other NGS-based proteomics platforms are based on probes that have been trained on monomeric recombinant proteins. They are limited to predicting the abundance of the canonical forms of these proteins. The Aptamarker platform does not rely on pre-trained probes. Through a combination of defining the fingerprint in the library of a recombinant protein and the fingerprint for the same protein as it exists in biological samples it is possible to define the proportion of a protein in different states across samples.

This capacity to characterize structural differences, PTMs, or isoforms across samples adds another dimension with the Aptamarker platform.

At the same time, we have also built in predictions of the canonical forms of a panel of cytokine proteins. As such end-users still obtain the information derived from other platforms and add additional dimensions of discovery capacity.

This is a common question.Outside of the existing cytokine abundance panel, the identity of the proteins bound by the Aptamarkers is not known. We see this as a necessary enablement of discovery.

If everything is defined, then it would not be possible to discover protein states that have not yet been discovered.

The identity of the proteins bound to Aptamarkers can be determined in many different ways.

• Hypothesis testing: In many cases prior to a study knowledge exists regarding existing biomarkers. Purified forms of these proteins can be applied to the library to identify their fingerprint. This is them used to track these biomarkers across samples. This is implicit to the platform and enables characterization of variance in a medical state that is due to known biomarkers and variance that is due to previously unknown biomarkers by Aptamarkers that are not related to known biomarkers.
• ELISA o SIMOA measurements: Another approach is to measure the known biomarkers across the same samples with an alternative approach such as an ELISA or SIMOA measurement. Then we identify those Aptamarkers that covary across samples with variance in this biomarker. The interesting data is the portion of variance that remains unexplained. This is presumably due to changes in the biomarker. Such information can then be use to improve the predictive potential for known biomarkers.
• Reverse proteomics: Where the target is truly unknown the Aptamarker can be immobilized on resin and exposed to the biological sample. The identity of the target that remains bound can be determined with mass spec analysis.

• 1) Traditional protein abundance predictions:

This data is similar to other NGS-based proteomics platforms, the data analysis code provides a list of 44 cytokine proteins and their predicted abundance (pg/mL). We are building other protein panels.

• 2)The frequency of all 428,686 structures across all samples.

Neoventures shares the code to convert NGS reads to structure reads. This data is then provided in bioinformatics ready analysis format as either frequencies or counts. Counts can be directly analyzed using existing DeSeq2 software.

These structural features provide the basis for at least three general types of analysis:

• -Volcano plots for enrichment and depletion of structures between groups of samples.
• -Cytoscape analysis of structural relationships between enriched Aptamarkers
• -Covariance analysis of structure frequencies across samples.

These applications result first in the identification of Aptamarker structures that are diagnostic for medical states, and then refinement of this data into sets of Aptamarkers that bind to different epitopes on the same proteins.