Gustavo de Los Campos

Gustavo de Los Campos

Auto-generated profile, based on publicly available information

Gustavo de Los Campos

Department of Epidemiology and Biostatistics, Michigan State University, East Lansing, USA

Research Focus & Metrics

Academic Career Length

14Years (2013-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
5 May 2015 in PLOS Genetics
Verified
Journal Article
Open access
PDF file
GCGustavo de Los Campos
Gustavo de Los Campos
DSDaniel Sorensen
Daniel Sorensen
DGDaniel Gianola
Daniel Gianola
Whole-genome regression methods are being increasingly used for the analysis and prediction of complex traits and diseases. In human genetics, these methods are commonly used for inferences about genetic parameters, such as the amount of genetic variance among individuals or the proportion of phenotypic variance that can be explained by regression on molecular markers. This is so even though some of the assumptions commonly adopted for data analysis are at odds with important quantitative genetic concepts. In this article we develop theory that leads to a precise definition of parameters arising in high dimensional genomic regressions; we focus on the so-called genomic heritability: the proportion of variance of a trait that can be explained (in the population) by a linear regression on a set of markers. We propose a definition of this parameter that is framed within the classical quantitative genetics theory and show that the genomic heritability and the trait heritability parameters are equal only when all causal variants are typed. Further, we discuss how the genomic variance and genomic heritability, defined as quantitative genetic parameters, relate to parameters of statistical models commonly used for inferences, and indicate potential inferential problems that are assessed further using simulations. When a large proportion of the markers used in the analysis are in LE with QTL the likelihood function can be misspecified. This can induce a sizable finite-sample bias and, possibly, lack of consistency of likelihood (or Bayesian) estimates. This situation can be encountered if the individuals in the sample are distantly related and linkage disequilibrium spans over short regions. This bias does not negate the use of whole-genome regression models as predictive machines; however, our results indicate that caution is needed when using marker-based regressions for inferences about population parameters such as the genomic heritability. Whole-genome regression (WGR) methods are being increasingly used for inferring the proportion of variance that can be explained by a linear regression on a massive number of markers, called ‘genomic heritability.’ However, the statistical assumptions involved in WGRs are somewhat at odds with important quantitative genetics concepts. We argue and show that the parameters of the statistical model used for data analysis typically bear a tenuous relationship with the quantitative genetic parameters of interest. We also study, using simulations, the extent of bias of likelihood-based estimates. We conclude that under certain circumstances estimates can have a sizable finite-sample bias; therefore, caution needs to be exercised when interpreting parameter estimates derived from WGR models.

1 November 2019 in G3: Genes, Genomes, Genetics
Verified
Journal Article
Open access
PDF file
CAChristina B. Azodi
Christina B. Azodi
EBEmily Bolger
Emily Bolger
AMAndrew McCarren
Andrew McCarren
MRMark Roantree
Mark Roantree
GCGustavo de Los Campos
Gustavo de Los Campos
SSShin-Han Shiu
Shin-Han Shiu
The usefulness of genomic prediction in crop and livestock breeding programs has prompted efforts to develop new and improved genomic prediction algorithms, such as artificial neural networks and gradient tree boosting. However, the performance of these algorithms has not been compared in a systematic manner using a wide range of datasets and models. Using data of 18 traits across six plant species with different marker densities and training population sizes, we compared the performance of six linear and six non-linear algorithms. First, we found that hyperparameter selection was necessary for all non-linear algorithms and that feature selection prior to model training was critical for artificial neural networks when the markers greatly outnumbered the number of training lines. Across all species and trait combinations, no one algorithm performed best, however predictions based on a combination of results from multiple algorithms (i.e., ensemble predictions) performed consistently well. While linear and non-linear algorithms performed best for a similar number of traits, the performance of non-linear algorithms vary more between traits. Although artificial neural networks did not perform best for any trait, we identified strategies (i.e., feature selection, seeded starting weights) that boosted their performance to near the level of other algorithms. Our results highlight the importance of algorithm selection for the prediction of trait values.

31 August 2018 in Genetics
Verified
Journal Article
Closed access
PDF file
PBPau Bellot
Pau Bellot
GCGustavo de Los Campos
Gustavo de Los Campos
MPMiguel Perez-Enciso
Miguel Perez-Enciso
The genetic analysis of complex traits does not escape the current excitement around artificial intelligence, including a renewed interest in “deep learning” (DL) techniques such as Multilayer Perceptrons (MLPs) and Convolutional Neural Networks (CNNs). However, the performance of DL for genomic prediction of complex human traits has not been comprehensively tested. To provide an evaluation of MLPs and CNNs, we used data from distantly related white Caucasian individuals (n ∼100k individuals, m ∼500k SNPs, and k = 1000) of the interim release of the UK Biobank. We analyzed a total of five phenotypes: height, bone heel mineral density, body mass index, systolic blood pressure, and waist–hip ratio, with genomic heritabilities ranging from ∼0.20 to 0.70. After hyperparameter optimization using a genetic algorithm, we considered several configurations, from shallow to deep learners, and compared the predictive performance of MLPs and CNNs with that of Bayesian linear regressions across sets of SNPs (from 10k to 50k) that were preselected using single-marker regression analyses. For height, a highly heritable phenotype, all methods performed similarly, although CNNs were slightly but consistently worse. For the rest of the phenotypes, the performance of some CNNs was comparable or slightly better than linear methods. Performance of MLPs was highly dependent on SNP set and phenotype. In all, over the range of traits evaluated in this study, CNN performance was competitive to linear models, but we did not find any case where DL outperformed the linear model by a sizable margin. We suggest that more research is needed to adapt CNN methodology, originally motivated by image analysis, to genetic-based problems in order for CNNs to be competitive with linear models.

22 October 2019 in The Plant Cell
Verified
Journal Article
Closed access
PDF file
CAChristina B. Azodi
Christina B. Azodi
JPJeremy Pardo
Jeremy Pardo
RVRobert VanBuren
Robert VanBuren
GCGustavo de Los Campos
Gustavo de Los Campos
SSShin-Han Shiu
Shin-Han Shiu
The ability to predict traits from genome-wide sequence information (i.e., genomic prediction) has improved our understanding of the genetic basis of complex traits and transformed breeding practices. Transcriptome data may also be useful for genomic prediction. However, it remains unclear how well transcript levels can predict traits, particularly when traits are scored at different development stages. Using maize (Zea mays) genetic markers and transcript levels from seedlings to predict mature plant traits, we found that transcript and genetic marker models have similar performance. When the transcripts and genetic markers with the greatest weights (i.e., the most important) in those models were used in one joint model, performance increased. Furthermore, genetic markers important for predictions were not close to or identified as regulatory variants for important transcripts. These findings demonstrate that transcript levels are useful for predicting traits and that their predictive power is not simply due to genetic variation in the transcribed genomic regions. Finally, genetic marker models identified only 1 of 14 benchmark flowering-time genes, while transcript models identified 5. These data highlight that, in addition to being useful for genomic prediction, transcriptome data can provide a link between traits and variation that cannot be readily captured at the sequence level.

27 August 2018 in Genetics
Verified
Journal Article
Closed access
PDF file
LLLouis Lello
Louis Lello
SASteven G. Avery
Steven G. Avery
LTLaurent Tellier
Laurent Tellier
AVAna I Vazquez
Ana I Vazquez
GCGustavo de Los Campos
Gustavo de Los Campos
stephen hsustephen hsu
stephen hsu
We construct genomic predictors for heritable but extremely complex human quantitative traits (height, heel bone density, and educational attainment) using modern methods in high dimensional statistics (i.e., machine learning). The constructed predictors explain, respectively, ∼40, 20, and 9% of total variance for the three traits, in data not used for training. For example, predicted heights correlate ∼0.65 with actual height; actual heights of most individuals in validation samples are within a few centimeters of the prediction. The proportion of variance explained for height is comparable to the estimated common SNP heritability from genome-wide complex trait analysis (GCTA), and seems to be close to its asymptotic value (i.e., as sample size goes to infinity), suggesting that we have captured most of the heritability for SNPs. Thus, our results close the gap between prediction R-squared and common SNP heritability. The ∼20k activated SNPs in our height predictor reveal the genetic architecture of human height, at least for common variants. Our primary dataset is the UK Biobank cohort, comprised of almost 500k individual genotypes with multiple phenotypes. We also use other datasets and SNPs found in earlier genome-wide association studies (GWAS) for out-of-sample validation of our results.

3 January 2017 in Plant Methods
Verified
Journal Article
Open access
PDF file
OMOsval Antonio Antonio Montesinos-López
Osval Antonio Antonio Montesinos-López
AMAbelardo Montesinos-López
Abelardo Montesinos-López
JCJose Crossa
Jose Crossa
GCGustavo de Los Campos
Gustavo de Los Campos
GAGregorio Alvarado
Gregorio Alvarado
JBJuan Burgueño
Juan Burgueño
Modern agriculture uses hyperspectral cameras to obtain hundreds of reflectance data measured at discrete narrow bands to cover the whole visible light spectrum and part of the infrared and ultraviolet light spectra, depending on the camera. This information is used to construct vegetation indices (VI) (e.g., green normalized difference vegetation index or GNDVI, simple ratio or SRa, etc.) which are used for the prediction of primary traits (e.g., biomass). However, these indices only use some bands and are cultivar-specific; therefore they lose considerable information and are not robust for all cultivars. This study proposes models that use all available bands as predictors to increase prediction accuracy; we compared these approaches with eight conventional vegetation indexes (VIs) constructed using only some bands. The data set we used comes from CIMMYT’s global wheat program and comprises 1170 genotypes evaluated for grain yield (ton/ha) in five environments (Drought, Irrigated, EarlyHeat, Melgas and Reduced Irrigated); the reflectance data were measured in 250 discrete narrow bands ranging between 392 and 851 nm. The proposed models for the simultaneous analysis of all the bands were ordinal least square (OLS), Bayes B, principal components with Bayes B, functional B-spline, functional Fourier and functional partial least square. The results of these models were compared with the OLS performed using as predictors each of the eight VIs individually and combined. We found that using all bands simultaneously increased prediction accuracy more than using VI alone. The Splines and Fourier models had the best prediction accuracy for each of the nine time-points under study. Combining image data collected at different time-points led to a small increase in prediction accuracy relative to models that use data from a single time-point. Also, using bands with heritabilities larger than 0.5 only in Drought as predictor variables showed improvements in prediction accuracy.