Content uploaded by Andre Altmann
Author content
All content in this area was uploaded by Andre Altmann on Nov 11, 2019
Content may be subject to copyright.
EVIDENCE FOR BIAS OF GENETIC ANCESTRY IN RESTING STATE FUNCTIONAL MRI
Andre Altmann?Janaina Mourao-Miranda?†
?Centre for Medical Image Computing
University College London, UK
†Max Planck University College London Centre for Computational Psychiatry and Ageing Research
University College London, UK
ABSTRACT
Resting state functional magnetic resonance imaging (rs-
fMRI) is a popular imaging modality for mapping the func-
tional connectivity of the brain. Rs-fMRI is, just like other
neuroimaging modalities, subject to a series of technical and
subject level biases that change the inferred connectivity pat-
tern. In this work we predicted genetic ancestry from rs-fMRI
connectivity data at very high performance (area under the
ROC curve of 0.93). Thereby, we demonstrated that genetic
ancestry is encoded in the functional connectivity pattern of
the brain at rest. Consequently, genetic ancestry constitutes a
bias that should be accounted for in the analysis of rs-fMRI
data.
Index Terms—resting state fMRI, genetics, ancestry,
imaging genetics, machine learning
1. INTRODUCTION
Task free or resting state (rs) functional magnetic resonance
imaging (fMRI) is an increasingly popular modality to map
the functional connectivity pattern of the brain. Recent rs-
fMRI applications include functional brain parcellation [1]
and disease biomarkers [2]. However, rs-fMRI is subject to
various biases that alter the inferred functional connectiv-
ity and may obscure true disease effects or genuine brain
function. For instance, observed connectivity patterns can
vary strongly with scanner model, pulse sequence and scan
site [3]. Also, involuntary head movement of the subject
during the scan can induce spurious functional connectiv-
ity between brain regions [4]. In addition to these technical
biases, there are also biases originating from subject level
characteristics: subjects’ sex and age have been reported to
exhibit wide-spread effects on the observed resting state con-
nectivity pattern [3]. In fact, subjects’ age can be successfully
predicted from rs-fMRI [5]. Furthermore, the involuntary
A.A. holds an Medical Research Council eMedLab Medical Bioinfor-
matics Career Development Fellowship. This work was supported by the
Medical Research Council [grant number MR/L016311/1]. JMM holds a
Wellcome Trust Senior Research Fellowship in Basic Biomedical Science
[grant number WT102845/Z/13/Z].
changes in cognitive states and vigilance levels has profound
changes in functional brain connectivity [6].
These biases are often taken into consideration when plan-
ning rs-fMRI studies. Typically, technical biases are mini-
mized by limiting studies to a single MRI scanner and by
addressing known artifacts, e.g., the ones arising from head
movement. Subject level characteristics are either considered
by matching the distributions, e.g., of age and sex, between
the studied disease groups or by including these subject char-
acteristics as confounding variables in the statistical analy-
sis. However, there is one additional subject level character-
istic that is known to affect head and brain morphology but is
rarely considered as a confound in rs-fMRI studies or brain
imaging studies in general: genetic ancestry.
The human genome was in part shaped by mankind’s mi-
gration history across the globe. Statistical analysis of genetic
data can reveal these ancient migration patterns. For instance,
the two main principal axes of variation in genetic similarity
of European individuals reflect the north-south and east-west
gradient within Europe [7]. The same principle holds true
for differences between continental regions across the world.
Thus, it is possible to reliable extract ancestry information
from genetic data, which is referred to as genetic ancestry and
mostly reflects subjects’ self-reported ethnicity.
Previous work on the relation of genetic ancestry and
brain imaging has demonstrated that head and brain morphol-
ogy of people with European ancestry follow the same north-
south and east-west pattern [8]. Follow-up work demon-
strated that the human cortical surface encodes the genetic
ancestry and that regional patterns of cortical folding and
gyrification are unique and complex for each continental
ancestry [9].
In this paper we demonstrate that functional connectiv-
ity networks obtained from high-quality rs-fMRI can reliably
predict genetic ancestry derived from genome wide genotyp-
ing data. In section 2 we describe the data used for this work
as well as the statistical analysis. Section 3 summarizes the
results which are discussed in section 4.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/440776doi: bioRxiv preprint first posted online Oct. 11, 2018;
2. MATERIAL AND METHODS
2.1. Genetic and imaging data
We obtained rs-fMRI data and matched genetic data from
the Young Adult study of the Human Connectome Project
(HCP) [10]. The HCP aims at charting the neural pathways
that underlie brain function and behavior and has acquired
high-quality neuroimaging data in over 1,100 healthy young
adults aged 22 - 35. Behavioral and other individual subject
measure data are available on all subjects.
2.1.1. Imaging data
The HCP acquired four imaging modalities using a 3T MR
for all subjects: structural images (T1w and T2w), resting-
state fMRI (rs-fMRI), task-fMRI (tfMRI), and high angular
resolution diffusion imaging (dMRI). The rs-fMRI data were
acquired in four runs of approximately 15 minutes each, two
runs in one session and two in another session, with eyes open
with relaxed fixation on a projected bright cross-hair on a dark
background (and presented in a darkened room). The TR was
0.75s amounting to 4,800 time-points per subject.
Detailed description of acquisition parameters and pro-
cessing steps can be found in [11]. In brief, group-average
parcellations for all subjects were obtained using group-ICA
at several different dimensionalities (15, 25, 50, 100, 200,
300). Next, subject-specific sets of node time series were
extracted, where each ICA-defined region of interest (ROI)
acted as a node. Then for each subject, a node ×node connec-
tivity matrix was created by computing the temporal correla-
tion between each pair of nodes. Correlation estimates were
based either on Pearson’s correlation or partial correlation us-
ing Tikhonov regularization. For the analyses presented here
we obtained the connectivity matrices based on partial corre-
lation at each of the six resolution levels.
The subjects’ matrices were vectorized to render them
amendable for machine learning; we extracted the upper right
triangle (omitting the diagonal values) of the correlation esti-
mates, thus generating for each subject feature vectors, x, of
length |x|=d2
−d
2, where dis the number of input ROIs.
2.1.2. Genetic data
Genetic data in HCP were available for 1,141 participants.
For this work we accessed the genome-wide genotyping data
that measured the identity of 2,119,803 genetic variants. In
brief, genotyping arrays assess the identity of genetic variants
at predefined locations in the genome. A variant can either
be identical to the one listed in the human reference genome
(reference) or differ from the reference (alternative). Humans
carry two copies of the genome in each cell. For analysis
purposes, one simply counts for each tested genetic variant
the number of alternative copies. Thus, at each of the two
million positions either 0, 1 or 2 is recorded.
2.2. Estimation of genetic ancestry
We used SNPweights [12] to obtain predictions for genetic
ancestry for four continental groups: Europeans (CEU),
African (YRI), Asian (ASI) and Native American (NAT).
SNPweights provides for each subject, i, and each continental
group, g, a probability pi
g∈[0,1]. Furthermore, probabilities
for each subject sum to 1 for all Nsubjects:
∀i∈[1, .., N ] : X
g
pi
g= 1.0.
2.3. Statistical learning
Our objective was to predict genetic ancestry from rs-fMRI
connectivity data. We employed the elastic net classifier [13],
which uses both the `1and `2penalties during training:
ˆ
β= argmin
β
L(y, ˆy) + λαkβk1+ (1 −α)kβk2
2,
where yis the vector of target values, βis a vector of model
coefficients for each feature, ˆy=f(X, β )is the vector of
model predictions, Xis the feature matrix, L(y, ˆy)is the
loss function, kβk1=P|βj|is the `1penalty, and kβk2=
qPβ2
jis the `2penalty. The loss function is typically either
the squared loss for regression problems or the logistic loss
for classification problems. The αparameter trades off the
`1and `2penalties. Of note, α= 0 corresponds to ridge re-
gression and α= 1 corresponds to the least absolute shrink-
age and selection operator (LASSO) regression. Thus, for
settings of αother than 0, the elastic net produces a sparse
solution with many entries in ˆ
βset to 0. In practice, the ex-
act choice of α(other than at the two extremes) has limited
impact on model performance. Therefore, we set a priory
α= 0.5. A more crucial parameter is λ, which trades off the
the amount of regularization with the model fit to the train-
ing data. Larger values of λtend to result in models with
more entries in βbeing set to 0 (i.e., sparser models). The λ
parameter is typically optimized using cross-validation (CV).
For the experiments described in this work we used the elastic
net implementation in the R package glmnet.
Depending on the selected resolution, each subjects’ fea-
ture vector ranged from 105 to 44,850 entries for 15 and 300
ROIs, respectively.
The target vector ywas based on the genetic ancestry pre-
diction. We dichotomized the quantitative output in order to
be able to perform classification rather than regression. One
consideration was that the dataset is rather imbalanced (see
section 3) and the target values ranged from 0.0 to 1.0 and are
therefore likely to suffer from the floor and ceiling effects. To
this end we employed two cutoffs for pCEU: 0.5 constituting
”mainly European” and 0.9 constituting ”predominantly Eu-
ropean”.
Classifier performance was assessed using receiver op-
erating characteristics (ROC) curves and the area under the
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/440776doi: bioRxiv preprint first posted online Oct. 11, 2018;
ROC curve (AUC). We used nested CV with 10 outer folds
and 5 inner folds. In nested CV, the data is first split into
outer folds (10 here). Then, data from all-but-one outer fold
are used to train the classifier. The classifier’s parameters (λ
here) are optimized through the inner CV (5-fold CV here).
The resulting classifier is then used to predict the label for
the data in the left-out outer fold to obtain an unbiased per-
formance estimate. The process is repeated until each outer
fold served once as hold-out data. Within the inner CV, we
selected the largest value of λ(i.e., the sparsest model) such
that the AUC was within 1 standard error of the maximum
AUC achieved by all tested λs; this strategy (”1 standard-error
rule”) is commonly employed to minimize the risk of over-
fitting by selecting a sparser model that does not perform sig-
nificantly worse than the best model [13]. Of note, the HCP
data contains data from siblings, which have a high genetic
similarity. In order to avoid confounding the performance es-
timate in cases where one sibling contributes to training and
the other sibling contributes to the testing data, we have sam-
pled the outer CV folds such that all family members were
placed within the same fold. In addition, the sampling for the
the outer fold was retained for all experiments to ensure that
results are comparable between settings.
In order to put the AUC results into the broader context
we also predicted subjects’ sex from the same rs-fMRI data
with the same CV folds using the 300 ROI resolution.
3. RESULTS
There were 1003 subjects with processed rs-fMRI data avail-
able through HCP, of these, 950 subjects (502 females) also
had genome-wide genotyping data available. SNPweights
successfully produced genetic ancestry predictions for the
four continental groups for all 950 subjects. When consid-
ering the highest probability for predicted genetic ancestry
(i.e., ”argmax”), then HCP comprises 764 CEU, 138 YRI,
39 ASI and 9 NAT subjects. Figure 1 depicts the distribution
of genetic ancestry predictions restricted to the three most
predominant groups in HCP: CEU,YRI and ASI. It becomes
evident that not only the corners of the triangle are populated,
i.e., representing uniform genetic ancestry, but there are also
many subjects with mixed ancestries, referred to as genetic
admixture.
As rationalized in section 2, we dichotomized the genetic
ancestry information for classification purposes. Given that
the majority of the HCP participants were of European ances-
try, we set a cutoff at pCEU >0.5resulting in 748 CEU and
202 non-CEU subjects and the other second more restrictive
cutoff at pCEU >0.9resulting in 651 CEU and 299 non-CEU
participants. We then employed the elastic net classifier to
predict CEU status using the vectorized rs-fMRI connectiv-
ity matrices. Practically, we were aiming to classify subjects
who are in the lower right corner of the triangle in Figure 1
versus subjects who are not, based on their rs-fMRI connec-
Fig. 1. Distribution of the three main genetic ancestries in the
HCP dataset. 80 subjects with pNAT >0.05 had been omitted
from this plot.
tivity. Table 1 summarizes the mean and standard deviation
of AUC values for the 10 outer folds of the CV for both ge-
netic ancestry cutoffs and all six levels of resolution. Fig-
ure 2 depicts representative ROC curves for selected ROI res-
olutions at the pCEU >0.5cutoff. AUC values ranged from
0.72 to 0.93; models using a finer-grained cortical parcellation
showed higher performance despite a high features to samples
ratio of ≈47 when using connectivity matrices based on 300
ROIs.
Subjects’ sex could also be predicted at very high ac-
curacy with mean AUC of 0.98 (and standard deviation of
0.016).
Table 1. Classifier performance quantified by mean AUC and
standard deviation from 10 ×5-fold nested CV for different
settings of pCEU cutoff and number of ROIs.
ROIs dim pCEU >0.5pCEU >0.9
15 105 0.78 (0.088) 0.72 (0.049)
25 300 0.81 (0.060) 0.76 (0.061)
50 1225 0.86 (0.055) 0.83 (0.063)
100 4950 0.91 (0.046) 0.86 (0.061)
200 19900 0.92 (0.039) 0.88 (0.055)
300 44850 0.93 (0.036) 0.87 (0.032)
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/440776doi: bioRxiv preprint first posted online Oct. 11, 2018;
Fig. 2. Averaged ROC curves from the 10 outer CV-folds.
Class labels were based on the pCEU >0.5cutoff. Dashed,
dotted and solid lines correspond to models based on 15, 50
and 300 ROIs, respectively. Whiskers indicate 1 standard er-
ror estimates.
4. DISCUSSION
This is the first work demonstrating that genetic ancestry is
highly predictable from rs-fMRI connectivity patterns. Our
results indicate that genetic ancestry is a serious bias that
modifies estimated brain connectivity and may mask genuine
differences or may introduce spurious differences in rs-fMRI
analyses between groups, e.g., a disease group and a control
group. The extent of the bias is not as pronounced as the in-
fluence of participants’ sex, which showed near perfect clas-
sification from rs-fMRI.
The exact origin of these apparent connectivity differ-
ences between continental ancestries remains elusive at the
moment. However, we hypothesize that the observed dif-
ferences are not based on true neuronal differences but that
they originate from differences in head and brain morphology
as reported in [8, 9]. These morphological differences may
be carried forward through the standard rs-fMRI processing
pipeline and affect the inferred functional connectivity. In
addition, rs-fMRI connectivity is based on correlations be-
tween blood-oxygen-level dependent (BOLD) signal time
series at rest. Thus, it is conceivable that genetic differences
contributing to blood circulation, perfusion and elasticity of
the vascular system may modify BOLD dynamics. This is ex-
emplified by reports identifying ethnicity as independent risk
factors for cardiovascular disease [14] and intracranial artery
tortuosity [15]. In addition, brain hemodynamic responses
are known to be heritable traits [16].
This study is not without limitations. Firstly, we chose
to dichotomize genetic ancestry and to perform classification
rather to train a regression model; this was mainly owing
to sample imbalance and to anticipated floor and ceiling ef-
fects when working with probability scores. Secondly, in this
proof-of-principle study we only classified European (CEU)
individuals from non-European individuals; studies on larger
datasets would be able to build classifiers for each continental
ancestry. Lastly, the analysis was limited to continental an-
cestries; with sufficiently large datasets it should be possible
to assess the effect of sub-ancestries within one continental
group on rs-fMRI inferred connectivity.
This work exemplifies that in the domain of rs-fMRI anal-
ysis there is a need to consider genetic ancestry as a confound
in the analysis. Given the previous findings of the strong
influence of genetic ancestry on regional volume and gyri-
fication [9], this consideration may extend to the entire neu-
roimaging field.
5. REFERENCES
[1] Matthew F Glasser, Timothy S Coalson, Emma C
Robinson, et al., “A multi-modal parcellation of hu-
man cerebral cortex,” Nature, vol. 536, no. 7615, pp.
171–178, Aug. 2016.
[2] Michael D Fox and Michael D Greicius, “Clinical appli-
cations of resting state functional connectivity.,” Fron-
tiers in systems neuroscience, vol. 4, pp. 19, 2010.
[3] Bharat B Biswal, Maarten Mennes, Xi-Nian Zuo, et al.,
“Toward discovery science of human brain function,”
Proceedings of the National Academy of Sciences of the
United States of America, vol. 107, no. 10, pp. 4734–
4739, Mar. 2010.
[4] Jonathan D Power, Anish Mitra, Timothy O Laumann,
et al., “Methods to detect, characterize, and remove mo-
tion artifact in resting state fMRI,” NeuroImage, vol. 84,
pp. 320–341, Jan. 2014.
[5] Nico U F Dosenbach, Binyam Nardos, Alexander L Co-
hen, et al., “Prediction of individual brain maturity using
fMRI,” Science (New York, NY), vol. 329, no. 5997, pp.
1358–1361, Sept. 2010.
[6] Andre Altmann, Manuel S Schr ¨
oter, Victor I Spoor-
maker, et al., “Validation of non-REM sleep stage de-
coding from resting state fMRI using linear support vec-
tor machines,” NeuroImage, vol. 125, pp. 544–555, Oct.
2015.
[7] John Novembre, Toby Johnson, Katarzyna Bryc, et al.,
“Genes mirror geography within Europe,” Nature, vol.
456, no. 7218, pp. 98–101, Nov. 2008.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/440776doi: bioRxiv preprint first posted online Oct. 11, 2018;
[8] Trygve E Bakken, Anders M Dale, and Nicholas J
Schork, “A geographic cline of skull and brain morphol-
ogy among individuals of European Ancestry,” Human
heredity, vol. 72, no. 1, pp. 35–44, 2011.
[9] Chun Chieh Fan, Hauke Bartsch, Andrew J Schork,
et al., “Modeling the 3D Geometry of the Cortical Sur-
face with Genetic Ancestry,” Current biology : CB, vol.
25, no. 15, pp. 1988–1992, Aug. 2015.
[10] David C Van Essen, Kamil Ugurbil, Edward J Auerbach,
et al., “The Human Connectome Project: a data acquisi-
tion perspective.,” NeuroImage, vol. 62, no. 4, pp. 2222–
2231, Oct. 2012.
[11] Stephen M Smith, Christian F Beckmann, Jesper Ander-
sson, et al., “Resting-state fMRI in the Human Connec-
tome Project.,” NeuroImage, vol. 80, pp. 144–168, Oct.
2013.
[12] Chia-Yen Chen, Samuela Pollack, David J Hunter, et al.,
“Improved ancestry inference using weights from exter-
nal reference panels,” Bioinformatics, vol. 29, no. 11,
pp. 1399–1406, June 2013.
[13] Jerome Friedman, Trevor Hastie, and Rob Tibshirani,
“Regularization Paths for Generalized Linear Models
via Coordinate Descent.,” Journal of Statistical Soft-
ware, vol. 33, no. 1, pp. 1–22, 2010.
[14] Sonia S Anand, Salim Yusuf, Vladimir Vuksan, et al.,
“Differences in risk factors, atherosclerosis, and cardio-
vascular disease between ethnic groups in Canada: the
Study of Health Assessment and Risk in Ethnic groups
(SHARE),” Lancet, vol. 356, no. 9226, pp. 279–284,
July 2000.
[15] Bum Joon Kim, Kyung Mi Lee, Sung-Ho Lee, et al.,
“Ethnic Differences in Intracranial Artery Tortuosity:
A Possible Reason for Different Locations of Cerebral
Atherosclerosis.,” Journal of stroke, vol. 20, no. 1, pp.
140–141, Jan. 2018.
[16] Zuyao Y Shan, Anna A E Vinkhuyzen, Paul M Thomp-
son, et al., “Genes influence the amplitude and timing
of brain hemodynamic responses,” NeuroImage, Sept.
2015.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/440776doi: bioRxiv preprint first posted online Oct. 11, 2018;
