Non-human Species

Non-human primates

SOMAmer reagents are generated to pure human proteins; due to the genetic similarity between primates it is hypothesized that the majority of SOMAmer reagents will cross react to non-human primate proteins and this is borne out in preliminary experiments. The concentrations of proteins are likely to be different in different sample types and species and this is reflected in the smaller sample volume requirements for non-human primates (110µL serum or EDTA-plasma) compared to 150 µL for humans (Table 2). The high evolutionary conservation between human and non-human primates suggests the SOMAscan assay will be a powerful proteomics tool for non-human primates.

Non-primates: Mice, rats, cats and dogs

SOMAmer reagents generated to pure human proteins have varying degrees of cross reactivity to non-human orthologs and therefore can be used to identify differential expression of some analytes in non-human samples. We have evaluated the cross reactivity of EDTA-plasma from mice, rats, cats and dogs, and our results suggest that the SOMAscan assay can detect and measure relative changes in hundreds of proteins reproducibly in small sample volumes from these species (1). Using the SOMAscan assay in new species has already led to novel discoveries in the mouse (2), indicating the SOMAscan assay can offer powerful, unbiased proteomics for many applications from preclinical models in drug discovery to animal health in veterinary sciences.

Sample volume requirements for EDTA-plasma and serum from tested species are listed in Table 2. We are continually investigating new sample types, such as cerebrospinal fluid from rats, and open to discussing developing additional sample types with collaborators. Specific protocols for various sample preparations are available (3). 

Species (serum/plasma) Requested volume (microliters)
Human 150
Mouse 70
Rat 50
Monkey 120
Dog 90
Cat 90

Mouse Models

Mouse-only models have been run in the SOMAscan assay with excellent results. In one study that profiled young and old mice, thirteen analytes were found which reliably distinguished young from old mice, p- value < 1.8e-5 (2). In total, 122 analytes were different between young and old mice with false discovery rate cutoff of 0.2. Many analytes had biological plausibility such as GDF-11 (2); for some the relevance to aging is novel at this time. In another mouse study, plasma from mice with genetic mutations linked to dysfunctional socialization was profiled in the SOMAscan assay. Seventeen proteins were found to be significantly different among all 6 groups with an FDR cutoff of 5% (q-value < 0.05). Out of all pair-wise comparisons 62 analytes were statistically significant, many of which were biologically relevant. In both studies, analytes reflective of sample handling were evident which permitted the exclusion of poorly handled samples, an important component of biomarker discovery efforts (4). 

The SOMAscan assay has also been used to detect differential expression in drug-treated, preclinical xenograft models. In one study, plasma was obtained from mice implanted with either a pancreatic or breast human xenograft before and after monoclonal antibody therapeutic treatment (unpublished data). Two hundred and fifty proteins changed in response to xenograft alone, likely a combination of proteins that come from the human xenograft and mouse response to xenograft. Many of these proteins decreased upon drug treatment correlating with decreasing tumor size. In another example tumor tissue from mouse xenografts were profiled in the SOMAscan assay, and while over 250 analytes changed significantly, these were human grafts and therefore the analytes are presumed to be human (5). These examples demonstrate the power of applying the “human” SOMAscan assay to other species, exposing deeper biological knowledge in preclinical models and leading to better health care for humans.

Profiling Plasma from Cats and Dogs

A small number of EDTA-plasma samples from both dog and cat were evaluated in the SOMAscan assay to estimate the number of signaling analytes. In this experiment, EDTA-plasma was obtained from 8 purebred beagles ranging from 3 – 5 years of age and eight different domestic short hair cats ranging between 8 – 22 months, both with equal gender representation. The samples were evaluated for signaling analytes and reproducibility by running nine replicates over three different assay runs. Signaling analytes were assessed by calculating the F-statistic (F-stat) of the ratio of population variance (among the 8 animals) to assay variance for each measurement. The population variance for “real” measurements must be greater than assay noise; non-signaling analytes are expected to have population variances similar to assay variance since both are measuring noise. The number of analytes found to be signaling in this limited population of animals was 294 for dogs and 687 for cats, was based on F-Stat > 29, a FDR corrected 95% confidence cutoff for 1129 measurements. The difference in the number of signaling analytes may reflect a larger natural proteomic variation in the outbred cats as compared to the 8 purebred beagles and may be expected to increase as more breeds are included. The %CV for the SOMAmer reagents binding to analytes in dog and cat plasma was excellent, with a median total CV of 3.9% and 4.4%, respectively. Only 17 and 15 analytes had a CV > 20%, for dogs and cats respectively. These results suggest the human-derived SOMAscan assay may have significant utility in preclinical development and veterinary sciences.


(1) Non-human species and the SOMAscan assay (Technical Note).
(2) Loffredo FS, et al. Growth Differentiation Factor 11 is a Circulating Factor that Reverses Age-Related Cardiomyopathy. (2013) Cell 153: 828-839.
(3) Recommended Sample Handling and Processing Guidelines (available on request from SomaLogic). 

(4) Williams S, et al. (2012). Exposing the criminal record of every blood sample: use of SOMAmer technology and sample mapping vectors to mitigate false biomarker discoveries. Poster presentation at Tri-Con 2012 in San Francisco.

(5) Ayers D, et al. (2012) Differential protein signatures in erlotinib-sensitive and resistant lung cancer cell lines in vitro and in vivo.  Poster presentation at ADAPT 2012 in Washington DC.