In Aptamers They Trust

Q: What is the clinical evidence from this study?

Study design: Editorial. Population: Cardiovascular disease and clinical phenotypes (Phecodes). Intervention: SOMAscan aptamer-based proteomic profiling and genetic imputation. Primary outcome: Associations between genetically imputed virtual proteome and EHR-derived clinical diagnoses (Phecodes).

Abhishek Joshi; Manuel Mayr

doi:10.1161/circulationaha.118.036823

November 27, 2018CirculationOpen Access

In Aptamers They Trust

View Full Paper

Ask AI

Bookmark

Share

Key Points

To examine the technical validity, scalability, and limitations of aptamer-based affinity assays and genetically imputed virtual proteomics in discovering disease biomarkers.
Expert commentary evaluating the SOMAscan v1.3 platform (1129 aptamers) used to identify single nucleotide polymorphism (SNP)-protein associations across the Framingham Offspring, KORA F4, and Malmö Diet and Cancer Study cohorts.
Assessed protein imputation across 41,288 individuals in the eMERGE cohort mapped against 1,128 electronic health record-derived PheCodes.
In evaluated discovery cohorts, gene-protein instruments predicted up to 60% of a protein's variance (median 8%), associating 55 virtual proteins with 89 clinical diagnoses.
Direct validation confirmed only 2 of 6 virtual proteins for atherosclerosis (CLC1B and PDGFR-beta) and 4 ApoE isoforms among 7 instruments for elevated LDL-C (median explained variation 3%).
Assay specificity concerns include non-specific binding, where 27.6% of SOMAmers (312/1129) generated signals in chicken plasma tenfold that of human plasma, and amino-acid substitutions altered target affinity in 32% of tested aptamers.

Structured PICO

P

Population

Editorial commentary discussing a study by Moseley et al. that evaluated the Framingham Offspring Cohort, KORA F4 dataset, Malmö Diet and Cancer Study (MDCS), and 41,288 genotyped individuals in the eMERGE cohort.

E

Exposure

SOMAscan v1.3 aptamer-based proteomic assay (measuring 1,129 plasma proteins) combined with SNP-based genetic imputation to generate a 'virtual proteome'

O

Outcome

Associations between genetically imputed plasma protein levels ('virtual proteome') and 1,128 clinical 'Phecodes' derived from Electronic Health Records (EHRs), with independent validation against elevated LDL-C and ultrasound-confirmed atherosclerosissurrogate

Limitations

Potential for non-specific binding (14% of aptamers in prior study) and cross-reactivity
Epitope effects where SNP-conferred amino-acid substitutions alter aptamer binding affinity (32% of panel)
Lack of internal controls and standard curves in SOMAscan assays, with potential batch and intra-plate effects
Reporting and coding biases inherent to Electronic Health Record (EHR) Phecodes
High false-negative rate and failure to detect known cardiovascular biomarkers like ApoB
SOMAscan uses a single binder and direct readout which can jeopardize specificity (e.g., 14% non-specific binding in the INTERVAL study)
Amino-acid substitutions conferred by SNPs significantly alter aptamer binding affinity in 32% of the panel, potentially causing artefactual abundance variations
Lack of internal controls and standard curves in SOMAscan, leading to potential batch or intra-plate normalization effects
EHR diagnosis codes (Phecodes) are collected for billing or incentivized care rather than research, introducing unquantified reporting biases
Well-known SNP-regulated cardiovascular biomarkers such as ApoB were missed by the analysis despite inclusion in the panel

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

While combining aptamer-based proteomics, genetic imputation, and EHR phenotyping enables high-throughput biomarker discovery, technical limitations of aptamer specificity, SNP-induced epitope effects, and EHR coding biases require orthogonal validation before clinical translation.

Abstract

lthough it has become affordable to sequence the entire human genome, measuring all proteins in a sample as complex as human plasma remains currently out of reach. The plasma proteome has an enormous dynamic range and variability, including splice variants, cleavage products, and posttranslational modifications. Antibody-based techniques have predominated, but other affinity-based techniques such as the SOMAscan aptamer assays from SomaLogic also promise the multiplexed measurement of proteins in a scalable manner.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Joshi et al. (2018) conducted an editorial in Cardiovascular disease and clinical phenotypes (Phecodes). SOMAscan aptamer-based proteomic profiling and genetic imputation was evaluated on Associations between genetically imputed virtual proteome and EHR-derived clinical diagnoses (Phecodes).

synapsesocial.com/papers/6a63d6d215f73d3e5bf28522 https://doi.org/10.1161/circulationaha.118.036823

View Full Paper

Ask AI

Bookmark

Share

In Aptamers They Trust

Key Points

Structured PICO

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

In Aptamers They Trust

Key Points

Structured PICO

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading