CheekAge: a next-generation buccal epigenetic aging clock associated with lifestyle and health.

Epigenetic aging clocks are computational models that predict age using DNA methylation information. Initially, first-generation clocks were developed to make predictions using CpGs that change with age. Over time, next-generation clocks were created using CpGs that relate to both age and health. Since existing next-generation clocks were constructed in blood, we sought to develop a next-generation clock optimized for prediction in cheek swabs, which are non-invasive and easy to collect. To do this, we collected MethylationEPIC data as well as lifestyle and health information from 8045 diverse adults. Using a novel simulated annealing approach that allowed us to incorporate lifestyle and health factors into training as well as a combination of CpG filtering, CpG clustering, and clock ensembling, we constructed CheekAge, an epigenetic aging clock that has a strong correlation with age, displays high test-retest reproducibility across replicates, and significantly associates with a plethora of lifestyle and health factors, such as BMI, smoking status, and alcohol intake. We validated CheekAge in an internal dataset and multiple publicly available datasets, including samples from patients with progeria or meningioma. In addition to exploring the underlying biology of the data and clock, we provide a free online tool that allows users to mine our methylomic data and predict epigenetic age.

Apigenin: a natural molecule at the intersection of sleep and aging.

NAD+, a pivotal coenzyme central to metabolism, exhibits a characteristic decline with age. In mice, NAD+ levels can be elevated via treatment with apigenin, a natural flavonoid that inhibits the NAD+-consuming glycoprotein CD38. In animal models, apigenin positively impacts both sleep and longevity. For example, apigenin improves learning and memory in older mice, reduces tumor proliferation in a mouse xenograft model of triple-negative breast cancer, and induces sedative effects in mice and rats. Moreover, apigenin elongates survival in fly models of neurodegenerative disease and apigenin glycosides increase lifespan in worms. Apigenin's therapeutic potential is underscored by human clinical studies using chamomile extract, which contains apigenin as an active ingredient. Collectively, chamomile extract has been reported to alleviate anxiety, improve mood, and relieve pain. Furthermore, dietary apigenin intake positively correlates with sleep quality in a large cohort of adults. Apigenin's electron-rich flavonoid structure gives it strong bonding capacity to diverse molecular structures across receptors and enzymes. The effects of apigenin extend beyond CD38 inhibition, encompassing agonistic and antagonistic modulation of various targets, including GABA and inflammatory pathways. Cumulatively, a large body of evidence positions apigenin as a unique molecule capable of influencing both aging and sleep. Further studies are warranted to better understand apigenin's nuanced mechanisms and clinical potential.

Glycine and aging: Evidence and mechanisms.

The restriction of calories, branched-chain amino acids, and methionine have all been shown to extend lifespan in model organisms. Recently, glycine was found to boost longevity in genetically heterogenous mice. This simple amino acid similarly extends lifespan in rats and improves health in mammalian models of age-related disease. While compelling data indicate that glycine is a pro-longevity molecule, divergent mechanisms may underlie its effects on aging. Glycine is abundant in collagen, a building block for glutathione, a precursor to creatine, and an acceptor for the enzyme glycine N-methyltransferase (GNMT). A review of the literature strongly implicates GNMT, which clears methionine from the body by taking a methyl group from S-adenosyl-L-methionine and methylating glycine to form sarcosine. In flies, Gnmt is required for reduced insulin/insulin-like growth factor 1 signaling and dietary restriction to fully extend lifespan. The geroprotector spermidine requires Gnmt to upregulate autophagy genes and boost longevity. Moreover, the overexpression of Gnmt is sufficient to extend lifespan and reduce methionine levels. Sarcosine, or methylglycine, declines with age in multiple species and is capable of inducing autophagy both in vitro and in vivo. Taken all together, existing evidence suggests that glycine prolongs life by mimicking methionine restriction and activating autophagy.

A set of common buccal CpGs that predict epigenetic age and associate with lifespan-regulating genes.

Epigenetic aging clocks are computational models that use DNA methylation sites to predict age. Since cheek swabs are non-invasive and painless, collecting DNA from buccal tissue is highly desirable. Here, we review 11 existing clocks that have been applied to buccal tissue. Two of these were exclusively trained on adults and, while moderately accurate, have not been used to capture health-relevant differences in epigenetic age. Using 130 common CpGs utilized by two or more existing buccal clocks, we generate a proof-of-concept predictor in an adult methylomic dataset. In addition to accurately estimating age (r = 0.95 and mean absolute error = 3.88 years), this clock predicted that Down syndrome subjects were significantly older relative to controls. A literature and database review of CpG-associated genes identified numerous genes (e.g., CLOCK, ELOVL2, and VGF) and molecules (e.g., alpha-linolenic acid, glycine, and spermidine) reported to influence lifespan and/or age-related disease in model organisms.

An integrative machine-learning meta-analysis of high-throughput omics data identifies age-specific hallmarks of Alzheimer's disease.

Alzheimer's disease (AD) is an incredibly complex and presently incurable age-related brain disorder. To better understand this debilitating disease, we collated and performed a meta-analysis on publicly available RNA-Seq, microarray, proteomics, and microRNA samples derived from AD patients and non-AD controls. 4089 samples originating from brain tissues and blood remained after applying quality filters. Since disease progression in AD correlates with age, we stratified this large dataset into three different age groups: < 75 years, 75-84 years, and ≥ 85 years. The RNA-Seq, microarray, and proteomics datasets were then combined into different integrated datasets. Ensemble machine learning was employed to identify genes and proteins that can accurately classify samples as either AD or control. These predictive inputs were then subjected to network-based enrichment analyses. The ability of genes/proteins associated with different pathways in the Molecular Signatures Database to diagnose AD was also tested. We separately identified microRNAs that can be used to make an AD diagnosis and subjected the predicted gene targets of the most predictive microRNAs to an enrichment analysis. The following key themes emerged from our machine learning and bioinformatics analyses: cell death, cellular senescence, energy metabolism, genomic integrity, glia, immune system, metal ion homeostasis, oxidative stress, proteostasis, and synaptic function. Many of the results demonstrated unique age-specificity. For example, terms highlighting cellular senescence only emerged in the earliest and intermediate age ranges while the majority of results relevant to cell death appeared in the youngest patients. Existing literature corroborates the importance of these hallmarks in AD.

Human age reversal: Fact or fiction?

Although chronological age correlates with various age-related diseases and conditions, it does not adequately reflect an individual's functional capacity, well-being, or mortality risk. In contrast, biological age provides information about overall health and indicates how rapidly or slowly a person is aging. Estimates of biological age are thought to be provided by aging clocks, which are computational models (e.g., elastic net) that use a set of inputs (e.g., DNA methylation sites) to make a prediction. In the past decade, aging clock studies have shown that several age-related diseases, social variables, and mental health conditions associate with an increase in predicted biological age relative to chronological age. This phenomenon of age acceleration is linked to a higher risk of premature mortality. More recent research has demonstrated that predicted biological age is sensitive to specific interventions. Human trials have reported that caloric restriction, a plant-based diet, lifestyle changes involving exercise, a drug regime including metformin, and vitamin D3 supplementation are all capable of slowing down or reversing an aging clock. Non-interventional studies have connected high-quality sleep, physical activity, a healthy diet, and other factors to age deceleration. Specific molecules have been associated with the reduction or reversal of predicted biological age, such as the antihypertensive drug doxazosin or the metabolite alpha-ketoglutarate. Although rigorous clinical trials are needed to validate these initial findings, existing data suggest that aging clocks are malleable in humans. Additional research is warranted to better understand these computational models and the clinical significance of lowering or reversing their outputs.

Pan-Tissue Aging Clock Genes That Have Intimate Connections with the Immune System and Age-Related Disease.

In our recent transcriptomic meta-analysis, we used random forest machine learning to accurately predict age in human blood, bone, brain, heart, and retina tissues given gene inputs. Although each tissue-specific model utilized a unique number of genes for age prediction, we found that the following six genes were prioritized in all five tissues: CHI3L2, CIDEC, FCGR3A, RPS4Y1, SLC11A1, and VTCN1. Since being selected for age prediction in multiple tissues is unique, we decided to explore these pan-tissue clock genes in greater detail. In the present study, we began by performing over-representation and network topology-based enrichment analyses in the Gene Ontology Biological Process database. These analyses revealed that the immunological terms "response to protozoan," "immune response," and "positive regulation of immune system process" were significantly enriched by these clock inputs. Expression analyses in mouse and human tissues identified that these inputs are frequently upregulated or downregulated with age. A detailed literature search showed that all six genes had noteworthy connections to age-related disease. For example, mice deficient in Cidec are protected against various metabolic defects, while suppressing VTCN1 inhibits age-related cancers in mouse models. Using a large multitissue transcriptomic dataset, we additionally generate a novel, minimalistic aging clock that can predict human age using just these six genes as inputs. Taken all together, these six genes are connected to diverse aspects of aging.

The protein inputs of an ultra-predictive aging clock represent viable anti-aging drug targets.

Machine learning models capable of predicting age given a set of inputs are referred to as aging clocks. We recently developed an aging clock that utilizes 491 plasma protein inputs, has an exceptional accuracy, and is capable of measuring biological age. Here, we demonstrate that this clock is extremely predictive (r = 0.95) when used to measure age in a novel plasma proteomic dataset derived from 370 human subjects aged 18-69 years. Over-representation analyses of the proteins that make up this clock in the Gene Ontology and Reactome databases predominantly implicated innate and adaptive immune system processes. Immunological drugs and various age-related diseases were enriched in the DrugBank and GLAD4U databases. By performing an extensive literature review, we find that at least 269 (54.8 %) of these inputs regulate lifespan and/or induce changes relevant to age-related disease when manipulated in an animal model. We also show that, in a large plasma proteomic dataset, the majority (57.2 %) of measurable clock proteins significantly change their expression level with human age. Different subsets of proteins were overlapped with distinct epigenetic, transcriptomic, and proteomic aging clocks. These findings indicate that the inputs of this age predictor likely represent a rich source of anti-aging drug targets.

Modeling the human aging transcriptome across tissues, health status, and sex.

Aging in humans is an incredibly complex biological process that leads to increased susceptibility to various diseases. Understanding which genes are associated with healthy aging can provide valuable insights into aging mechanisms and possible avenues for therapeutics to prolong healthy life. However, modeling this complex biological process requires an enormous collection of high-quality data along with cutting-edge computational methods. Here, we have compiled a large meta-analysis of gene expression data from RNA-Seq experiments available from the Sequence Read Archive. We began by reprocessing more than 6000 raw samples-including mapping, filtering, normalization, and batch correction-to generate 3060 high-quality samples spanning a large age range and multiple different tissues. We then used standard differential expression analyses and machine learning approaches to model and predict aging across the dataset, achieving an R2 value of 0.96 and a root-mean-square error of 3.22 years. These models allow us to explore aging across health status, sex, and tissue and provide novel insights into possible aging processes. We also explore how preprocessing parameters affect predictions and highlight the reproducibility limits of these machine learning models. Finally, we develop an online tool for predicting the ages of human transcriptomic samples given raw gene expression counts. Together, this study provides valuable resources and insights into the transcriptomics of human aging.

Data mining of human plasma proteins generates a multitude of highly predictive aging clocks that reflect different aspects of aging.

We previously identified 529 proteins that had been reported by multiple different studies to change their expression level with age in human plasma. In the present study, we measured the q-value and age coefficient of these proteins in a plasma proteomic dataset derived from 4263 individuals. A bioinformatics enrichment analysis of proteins that significantly trend toward increased expression with age strongly implicated diverse inflammatory processes. A literature search revealed that at least 64 of these 529 proteins are capable of regulating life span in an animal model. Nine of these proteins (AKT2, GDF11, GDF15, GHR, NAMPT, PAPPA, PLAU, PTEN, and SHC1) significantly extend life span when manipulated in mice or fish. By performing machine-learning modeling in a plasma proteomic dataset derived from 3301 individuals, we discover an ultra-predictive aging clock comprised of 491 protein entries. The Pearson correlation for this clock was 0.98 in the learning set and 0.96 in the test set while the median absolute error was 1.84 years in the learning set and 2.44 years in the test set. Using this clock, we demonstrate that aerobic-exercised trained individuals have a younger predicted age than physically sedentary subjects. By testing clocks associated with 1565 different Reactome pathways, we also show that proteins associated with signal transduction or the immune system are especially capable of predicting human age. We additionally generate a multitude of age predictors that reflect different aspects of aging. For example, a clock comprised of proteins that regulate life span in animal models accurately predicts age.

Stirred suspension bioreactors maintain naïve pluripotency of human pluripotent stem cells.

Due to their ability to standardize key physiological parameters, stirred suspension bioreactors can potentially scale the production of quality-controlled pluripotent stem cells (PSCs) for cell therapy application. Because of differences in bioreactor expansion efficiency between mouse (m) and human (h) PSCs, we investigated if conversion of hPSCs, from the conventional "primed" pluripotent state towards the "naïve" state prevalent in mPSCs, could be used to enhance hPSC production. Through transcriptomic enrichment of mechano-sensing signaling, the expression of epigenetic regulators, metabolomics, and cell-surface protein marker analyses, we show that the stirred suspension bioreactor environment helps maintain a naïve-like pluripotent state. Our research corroborates that converting hPSCs towards a naïve state enhances hPSC manufacturing and indicates a potentially important role for the stirred suspension bioreactor's mechanical environment in maintaining naïve-like pluripotency.

Doxorubicin generates senescent microglia that exhibit altered proteomes, higher levels of cytokine secretion, and a decreased ability to internalize amyloid ß.

Cellular senescence is defined by irreversible cell-cycle arrest and is an evolutionarily conserved hallmark of aging. In this study, we generate senescent microglial cells via exposure to the chemotherapy drug doxorubicin. Compared to control cells, doxorubicin-treated microglia exhibited an altered morphology characterized by an enlarged cell size, a flattened appearance, and the development of prominent filaments. Senescent cells harbored elevated levels of senescence associated-ß-galactosidase, p16Ink4a, and γ-H2AX. Senescent microglia were also less efficient at internalizing amyloid ß and pHrodo bioparticles. A detailed proteomic analysis using SWATH-MS identified 201 proteins that were significantly downregulated and 127 that were significantly upregulated in doxorubicin-treated microglia. Proteins involved in processes such as protein synthesis, RNA damage and repair, and protein degradation were largely downregulated while those compromising the integrity of the cell were predominantly upregulated. Various proteins involved in proteasomal processing were among the most significantly downregulated in senescent cells. Relevant to the deleterious senescence-associated secretory phenotype, senescent cells secreted higher levels of the inflammatory cytokines IL-6, IL-8, TNF-α, and GRO-α. Our data suggest that symptoms of brain aging and age-related neurodegenerative disease may be partially caused by defective phagocytosis, impaired proteasomal processing, and elevated cytokine secretion of senescent microglia.

Molecular analyses of glioblastoma stem-like cells and glioblastoma tissue.

Glioblastoma is a common, malignant brain tumor whose disease incidence increases with age. Glioblastoma stem-like cells (GSCs) are thought to contribute to cancer therapy resistance and to be responsible for tumor initiation, maintenance, and recurrence. This study utilizes both SNP array and gene expression profiling to better understand GSCs and their relation to malignant disease. Peripheral blood and primary glioblastoma tumor tissue were obtained from patients, the latter of which was used to generate GSCs as well as a CD133pos./CD15pos. subpopulation. The stem cell features of GSCs were confirmed via the immunofluorescent expression of Nestin, SOX2, and CD133. Both tumor tissue and the isolated primary cells shared unique abnormal genomic characteristics, including a gain of chromosome 7 as well as either a partial or complete loss of chromosome 10. Individual genomic differences were also observed, including the loss of chromosome 4 and segmental uniparental disomy of 9p24.3→p21.3 in GSCs. Gene expression profiling revealed 418 genes upregulated in tumor tissue vs. CD133pos./CD15pos. cells and 44 genes upregulated in CD133pos./CD15pos. cells vs. tumor tissue. Pathway analyses demonstrated that upregulated genes in CD133pos./CD15pos. cells are relevant to cell cycle processes and cancerogenesis. In summary, we detected previously undescribed genomic and gene expression differences when comparing tumor tissue and isolated stem-like subpopulations.

Systematic review and analysis of human proteomics aging studies unveils a novel proteomic aging clock and identifies key processes that change with age.

The development of clinical interventions that significantly improve human healthspan requires robust markers of biological age as well as thoughtful therapeutic targets. To promote these goals, we performed a systematic review and analysis of human aging and proteomics studies. The systematic review includes 36 different proteomics analyses, each of which identified proteins that significantly changed with age. We discovered 1,128 proteins that had been reported by at least two or more analyses and 32 proteins that had been reported by five or more analyses. Each of these 32 proteins has known connections relevant to aging and age-related disease. GDF15, for example, extends both lifespan and healthspan when overexpressed in mice and is additionally required for the anti-diabetic drug metformin to exert beneficial effects on body weight and energy balance. Bioinformatic enrichment analyses of our 1,128 commonly identified proteins heavily implicated processes relevant to inflammation, the extracellular matrix, and gene regulation. We additionally propose a novel proteomic aging clock comprised of proteins that were reported to change with age in plasma in three or more different studies. Using a large patient cohort comprised of 3,301 subjects (aged 18-76 years), we demonstrate that this clock is able to accurately predict human age.

Lipid Hydrolase Enzymes: Pragmatic Prolongevity Targets for Improved Human Healthspan?

Compelling evidence suggests that lipid metabolism, which plays critical roles in fat storage, cell membrane maintenance, and cell signaling, is intricately linked to aging. Lipid hydrolases are important enzymes that catalyze the hydrolysis of more complex lipids into simpler lipids. Diverse interventions targeting lipid hydrolases can prolong or shorten life in model organisms. For example, the genetic removal of or RNAi knockdown against a phospholipase can reduce lifespan in Caenorhabditis elegans, Drosophila melanogaster, and Mus musculus. The removal of lysosomal acid lipase results in premature death in mice, while its overexpression in nematodes generates lean, long-lived individuals. The overexpression or inhibition of diacylglycerol lipase leads to enhanced or reduced longevity, respectively, in both worms and flies. Lifespan can also be extended by knocking down triacylglycerol lipases in yeast, overexpressing fatty acid amide hydrolase in worms, or removing hepatic lipase in a mouse model of coronary disease. Conversely, flies lacking the triacylglycerol lipase Brummer are obese and short lived. Linking sphingolipids and aging, removing the sphingomyelinase inositol phosphosphingolipid phospholipase shortens chronological lifespan in Saccharomyces cerevisiae, while inhibiting an acid sphingomyelinase in worms or inactivating alkaline ceramidase in flies extends lifespan. The clinical potential of manipulating these enzymes is highlighted by the FDA-approved obesity drug orlistat, which is an inhibitor of pancreatic and hepatic lipases that induces weight loss and improves insulin/glucose homeostasis. Additional research is warranted to better understand how these lipid hydrolases impact aging and to determine if clinical interventions targeting them are capable of improving human healthspan.

The role of lipid metabolism in aging, lifespan regulation, and age-related disease.

An emerging body of data suggests that lipid metabolism has an important role to play in the aging process. Indeed, a plethora of dietary, pharmacological, genetic, and surgical lipid-related interventions extend lifespan in nematodes, fruit flies, mice, and rats. For example, the impairment of genes involved in ceramide and sphingolipid synthesis extends lifespan in both worms and flies. The overexpression of fatty acid amide hydrolase or lysosomal lipase prolongs life in Caenorhabditis elegans, while the overexpression of diacylglycerol lipase enhances longevity in both C. elegans and Drosophila melanogaster. The surgical removal of adipose tissue extends lifespan in rats, and increased expression of apolipoprotein D enhances survival in both flies and mice. Mouse lifespan can be additionally extended by the genetic deletion of diacylglycerol acyltransferase 1, treatment with the steroid 17-α-estradiol, or a ketogenic diet. Moreover, deletion of the phospholipase A2 receptor improves various healthspan parameters in a progeria mouse model. Genome-wide association studies have found several lipid-related variants to be associated with human aging. For example, the epsilon 2 and epsilon 4 alleles of apolipoprotein E are associated with extreme longevity and late-onset neurodegenerative disease, respectively. In humans, blood triglyceride levels tend to increase, while blood lysophosphatidylcholine levels tend to decrease with age. Specific sphingolipid and phospholipid blood profiles have also been shown to change with age and are associated with exceptional human longevity. These data suggest that lipid-related interventions may improve human healthspan and that blood lipids likely represent a rich source of human aging biomarkers.

Revamping the evolutionary theories of aging.

Radical lifespan disparities exist in the animal kingdom. While the ocean quahog can survive for half a millennium, the mayfly survives for less than 48 h. The evolutionary theories of aging seek to explain why such stark longevity differences exist and why a deleterious process like aging evolved. The classical mutation accumulation, antagonistic pleiotropy, and disposable soma theories predict that increased extrinsic mortality should select for the evolution of shorter lifespans and vice versa. Most experimental and comparative field studies conform to this prediction. Indeed, animals with extreme longevity (e.g., Greenland shark, bowhead whale, giant tortoise, vestimentiferan tubeworms) typically experience minimal predation. However, data from guppies, nematodes, and computational models show that increased extrinsic mortality can sometimes lead to longer evolved lifespans. The existence of theoretically immortal animals that experience extrinsic mortality - like planarian flatworms, panther worms, and hydra - further challenges classical assumptions. Octopuses pose another puzzle by exhibiting short lifespans and an uncanny intelligence, the latter of which is often associated with longevity and reduced extrinsic mortality. The evolutionary response to extrinsic mortality is likely dependent on multiple interacting factors in the organism, population, and ecology, including food availability, population density, reproductive cost, age-mortality interactions, and the mortality source.

Concise Review: Molecular Cytogenetics and Quality Control: Clinical Guardians for Pluripotent Stem Cells.

Now that induced pluripotent stem cell (iPSC)-based transplants have been performed in humans and organizations have begun producing clinical-grade iPSCs, it is imperative that strict quality control standards are agreed upon. This is essential as both ESCs and iPSCs have been shown to accumulate genomic aberrations during long-term culturing. These aberrations can include copy number variations, trisomy, amplifications of chromosomal regions, deletions of chromosomal regions, loss of heterozygosity, and epigenetic abnormalities. Moreover, although the differences between iPSCs and ESCs appear largely negligible when a high enough n number is used for comparison, the reprogramming process can generate further aberrations in iPSCs, including copy number variations and deletions in tumor-suppressor genes. If mutations or epigenetic signatures are present in parental cells, these can also be carried over into iPSCs. To maximize patient safety, we recommend a set of standards to be utilized when preparing iPSCs for clinical use. Reprogramming methods that do not involve genomic integration should be used. Cultured cells should be grown using feeder-free and serum-free systems to avoid animal contamination. Karyotyping, whole-genome sequencing, gene expression analyses, and standard sterility tests should all become routine quality control tests. Analysis of mitochondrial DNA integrity, whole-epigenome analyses, as well as single-cell genome sequencing of large cell populations may also prove beneficial. Furthermore, clinical-grade stem cells need to be produced under accepted regulatory good manufacturing process standards. The creation of haplobanks that provide major histocompatibility complex matching is also recommended to improve allogeneic stem cell engraftment. Stem Cells Translational Medicine 2018;7:867-875.

Mutant Best1 Expression and Impaired Phagocytosis in an iPSC Model of Autosomal Recessive Bestrophinopathy.

Autosomal recessive bestrophinopathy (ARB) is caused by mutations in the gene BEST1 which encodes bestrophin 1 (Best1), an anion channel expressed in retinal pigment epithelial (RPE) cells. It has been hypothesized that ARB represents the human null phenotype for BEST1 and that this occurs due to nonsense mediated decay (NMD). To test this hypothesis, we generated induced pluripotent stem cells (iPSCs) from a patient with ARB and her parents. After differentiation to retinal pigment epithelial (iPSC-RPE) cells, both BEST1 mRNA and Best1 protein expression were compared to controls. BEST1 mRNA expression levels, determined by quantitative PCR, were similar in ARB iPSC-RPE, parental cells, and genetically unrelated controls. Western blotting revealed that CRALBP and RPE65 were expressed within the range delineated by unrelated controls in iPSC-RPE from the ARB donor and her parents. Best1 protein was detected in different clones of ARB iPSC-RPE, but at reduced levels compared to all controls. When tested for the ability to phagocytose photoreceptor outer segments, ARB iPSC-RPE exhibited impaired internalization. These data suggest that impaired phagocytosis is a trait common to the bestrophinopathies. Furthermore, ARB is not universally the result of NMD and ARB, in this patient, is not due to the absence of Best1.

Disease modeling studies using induced pluripotent stem cells: are we using enough controls?

The comparison of differentiated induced pluripotent stem cells (iPSCs) derived from patients with disease to differentiated iPSCs derived from healthy patients enables powerful disease modeling. By performing an informal retrospective survey of disease modeling studies published in high impact journals, we found that the median and average number of controls used in these studies were 1 and 1.6, respectively. The bulk of these studies did not control for age, gender and ethnicity. Since a large proportion of phenotypic differences observed between iPSC lines are due to genetic variation or variation between lines, this is an insufficient number of controls to confidently rule out standard variation. Future studies need to include more controls and ensure that these controls are appropriately matched for gender, age and ethnicity.