3D Facial Norms Database

Although ample evidence exists that facial appearance and structure are highly heritable, there is a dearth of information regarding how variation in specific genes relates to the diversity of facial forms evident in our species. With the advent of affordable, non-invasive 3D surface imaging technology, it is now possible to capture detailed quantitative information about the face in a large number of individuals. By coupling state- of-the-art 3D imaging with advances in high-throughput genotyping, an unparalleled opportunity exists to map the genetic determinants of normal facial variation. An improved understanding of the relationship between genotype and facial phenotype may help illuminate the factors influencing liability to common craniofacial anomalies, particularly orofacial clefts, which are among the most prevalent birth defects in humans. This proposal has two major goals: (1) to construct a nonnative repository of 3D facial and genetic data and (2) to utilize this data repository to identify genes that influence normal midfacial variation. The first goal will focus on data generation and resource development and will involve the collection of 3D facial surface images and DNA samples on 3500 healthy Caucasian individuals (age 5-40) drawn from the general population. Quantitative facial measures will be extracted from the 3D images and all DNA samples will be genotyped for genome-wide SNP markers. Working in conjunction with the FaceBase hub, our intent is to create a scalable. Interactive and minable data resource available to outside investigators, which will contain facial measures, 3D images and genotypes. Ultimately, it is hoped that such a database will facilitate novel research initiatives. To illustrate this potential, the second goal of this proposal will focus on identifying SNPs associated with variation in midfacial morphology, including those facial features relevant to orofacial cleft predisposition. Salient measures of midfacial morphology will be derived from 3D facial surface images, and a genome-wide association approach will then be employed to identify polymorphisms that influence quantitative variation in the facial features of interest.


3D Facial Norms Database

Data is now available.  Please contact with your request to receive further information.

Oral Clefts: Moving from Genome-Wide Studies Toward Functional Genomics

We will follow up on signals from a genome wide association study (GWAS) of oral clefts now being conducted with support from U01-DE-004425; "International Consortium to Identify Genes & Interactions Controlling Oral Clefts", 2007-2009; TH Beaty, PI). Oral clefts are among the most common human birth defects and have a complex and heterogeneous etiology. Genotyping for this project should be completed in early 2009, and our analysis will identify genes influencing risk directly, those acting only in the presence of an environmental risk factor, and/or genes showing measurable parent-of-origin effects which may represent imprinting. In this response to the FaceBase initiative (RFA-DE-09-003), we will build upon our GWAS results by using high throughput sequencing (HTS) techniques on genes/regions yielding statistical evidence of linkage and disequilibrium in the GWAS. We will first focus on genes identified through analysis of single nucleotide polymorphic (SNP) markers from our GWAS and will use HTS to identify all variants (rare mutations and novel markers) that may be causal, directly or indirectly. We will then undertake a systematic analysis of intensity data and the 60,000 markers in regions of known copy number variants (CNV) available on this platform to test genes that may influence risk through structural variation.


Data is now available consisting of imputed markers generated with BEAGLE in CL/P case-parent trios from an international consortium. Imputation was performed taking family structure into account. For each population (Asian and European) ped- and map-formatted files ( are provided. 


Please submit a Data Access Request form if you are interested.  View Data Dictionary description.

Identification of miRNAs Involved in Midfacial Development and Clefting

The goal of this project is to identify microRNAs (miRNAs) involved in vertebrate mid-face development and to determine their function in this process. Our hypothesis is that a limited number of miRNAs play crucial roles in the patterning, proliferation and differentiation of the mid-face through precise spatial and temporal expression of specific protein-coding genes. Further, we hypothesize that both loss- and gain-of-function of these miRNAs leads to mid-facial and orofacial defects/clefting. In humans, defects in mid-facial development, including cleft lip/palate, account for the largest number of birth defects annually. Alone, over 400 syndromes result in cleft lip and palate, which has an occurrence rate of 1 in 800 live births. Our laboratories have a long-standing interest in determining the gene regulatory networks that control normal orofacial development, many of which are causative in animal models of mid-facial clefting. In our Preliminary Data, we show that microarray technology can generate comprehensive miRNA expression profiles and that bioinformatic analysis can define miRNAs whose expression patterns are conserved across vertebrates. In this proposal, we will investigate the identity and function of miRNAs expressed in the developing mid-face in three Specific Aims. In Aim 1, we will determine and validate the temporal and spatial expression patterns of miRNAs in the developing mouse maxillary/frontonasal prominences using miRNA microarray technology. In Aim 2, we will determine the in situ expression patterns of identified miRNAs in both mouse and zebrafish embryos to define those that show a conserved pattern of expression. In Aim 3, we will use zebrafish to determine the regulatory function and morphogenetic mode of action of the miRNAs by loss- and gain-of-function analysis. We propose using both mouse and zebrafish systems because their complementary well-established experimental and genetic methods make a detailed analysis of this process feasible. By defining miRNAs involved in mouse mid-facial development and then quickly and efficiently determining miRNA function in zebrafish, we will be able to elucidate the miRNAs involved in vertebrate mid-facial development. This is crucial to further our understanding of genetic events leading to human mid-facial birth defect syndromes.

Research on Functional Genomics, Image Analysis and Rescue of Cleft Palate

The development of the Face Base Consortium calls for a comprehensive research collaboration to facilitate data collection, organization, and optimized utilization of new and existing data on mid-facial development and malformations. Our laboratory has a long history of investigating the molecular and cellular mechanism of cleft palate. We have developed a proposal that builds on our strength and will focus on genomic and imaging analysis of selected and highly clinically relevant cleft palate animal models. Specifically, we will use the Tgfb, Tgfbr, Smad4, Msx1 and Fgfr2 mutant animal models that represent complete and sub-mucous cleft palate defects in humans as our entry point. Taking advantage of these animal models, we will work closely with several scientists to address the regulatory mechanism of CNC cell fate determination. Specifically, working with Dr. Marianne Bronner-Fraser at California Institute of Technology (Caltech), we will investigate whether the neural crest gene regulatory network of traditional vertebrate models is conserved and may exert its regulatory function during palatogenesis. In collaboration with Dr. Joseph Hacia at USC, we will discover critical components of the Tgf-b signaling network that are specifically involved in regulating the fate of CNC cells during palatogenesis. Working with Dr. Scott Fraser at Caltech, we will generate comprehensive and dynamic three-dimensional images of palatogenesis and malformations using microMRI and microCT. Finally, we have developed a strategy to screen for specific points of intervention within the gene regulatory network that will allow us to develop therapeutic strategies to prevent and rescue cleft palate. Our collective effort will not only generate tremendous resources for the Face Base Consortium but will also offer opportunities for extensive collaborations for future translational research on craniofacial birth defects.

Functional Analysis of Neural Crest and Palate: Imaging Craniofacial Development

The research plan will refine and deploy a set of advanced tools for the imaging of tissue structure, gene expression domains and cellular dynamics during craniofacial development. Volumetric imaging tools will be used to create accurate 3D atlases that can be digitally dissected to permit the tissue interactions and cellular events to be better understood in the forming faces of normal, mutant and perturbed mouse and avian embryos. Molecular imaging agents, optimized for imaging intact tissues, will be employed to create atlases of the molecular correlates of these embryos. Finally, intravital imaging tools will be used to study cell and tissue interactions as they take place, offering a view into the dynamic events that execute craniofacial development. The data will be assembled into atlases that will offer unprecedented tools for exploring the cellular, tissue and molecular correlates of craniofacial development. These atlases will be made widely available for the use of others in the Face Base consortium and the broader community of craniofacial researchers through an online resource created locally. In addition, they will be linked to other Face Base resources through established hubs, so that the data in our atlases can be used in concert with the molecular and structural data provided by other members of the consortium. The data acquisition pipeline established here in close collaboration with the laboratories of Dr. Yang Chai (University of Southern California) and Dr. Marianne Bronner-Fraser (Caltech) will create a model that can be expanded to phenotype and analyze other experimental systems.

Genetic Determinants of Orofacial Shape and Relationship to Cleft Lip/Palate

Orofacial clefts, principally cleft lip (CL), cleft palate (CP), and cleft lip and palate (CLP), are among the most common major birth defects, occurring in ~1/700 to 1/1000 live births in various populations around the world, ~70% as a sporadic, isolated abnormality. Such "non-syndromic" orofacial clefts act as complex traits, involving multiple genes and environmental risk factors. To date traditional genetic mapping approaches have identified only a few major susceptibility genes for non-syndromic orofacial clefts with certainty. Therefore, new approaches are clearly required. There is considerable evidence that orofacial malformations can occur at the extremes of the normal ranges of phenotypic variation of midfacial size and shape. Here we propose a novel approach to identify genes that regulate midfacial shape in mouse and human. We hypothesize that genes that are major contributors to normal orofacial size and shape will also have important roles in the occurrence of orofacial clefts. To identify such genes, we will perform detailed morphometric analysis of midfacial shape differences in informative mouse strains as well as in select human populations, combining these studies with genetic analyses to identify genes that control major determinants of midfacial morphometries. Our studies have shown that specific inbred strains of mice have heritable differences in measurable parameters of facial shape. We will take advantage of a valuable new resource we have developed, the mouse "Collaborative Cross" (CC), to con-elate heritable differences in facial shape among the 8 founder strains of the CC, along with select Recombinant Inbred lines and Recombinant Intercross (RIX), with detailed genetic mapping data for these mice. This approach will enable identification of quantitative trait loci (QTLs) that underlie these morphometric differences. We will complement our mouse studies with a similar analysis of humans, studying specific populations with different susceptibilities to orofacial clefts. These comparative studies will allow us to identify genes that underlie midfacial shape in humans. Together, these studies should provide a basis for understanding the relationship between human facial morphogenesis and susceptibility to orofacial clefts, and for initiating studies of the functions of these genes in animal models relevant to human orofacial development.


Data is now available.  Please contact with your request to receive further information.

Genome-Wide Atlas of Craniofacial Transcriptional Enhancers

Clefts of the lip and palate are among the most common craniofacial birth defects. They can co-occur with other symptoms as part of Mendelian disorders, but the majority of cases are non-syndromic and have a complex etiology. In some cases disrupted protein-coding genes have been identified as contributors to orofacial clefting risk. However, accumulating evidence from genome-wide association studies (GWAS) indicates that sequence variation in non-coding regions also strongly contributes to a variety of clinical disorders including orofacial clefting. While it is speculated that many of these variants affect disease through impacting on functional properties of distant-acting transcriptional enhancers, only very few isolated examples of such regulatory variation have been identified. This is likely due to the fact that the genomic location and function of the vast majority of distant-acting enhancers in the human genome remains unknown. To address the pressing need to identify on a genomic scale enhancers that are involved in face and palate development and likely relevant for clefting etiology, we propose here an integrated genomic and transgenic mouse strategy to identify craniofacial enhancers and characterize their activities. Specifically, we will use a ChlP-seq approach to identify genome-wide sets of enhancers that are active in mouse face and palate tissues at embryonic stages that are relevant for orofacial clefting. We will use a transgenic mouse enhancer screen to validate and characterize 130 of these enhancer predictions in detail by determining their in vivo activity patterns. Furthermore, we will identify disease-associated GWAS variants that map to craniofacial enhancers that we will have discovered. We will then test and compare the variant and normal sequences in the transgenic enhancer assay for differences in their in vivo activities. All of the genomic and in vivo datasets, as well as molecular reagents developed through these experiments will be made available to other investigators through the Face Base program in order to maximize their availability and accelerate the progress of biomedical and clinical studies of mid-face and palate development and orofacial clefting.

Global Gene Expression Atlas of Craniofacial Development

The objective is to create a global gene expression atlas of craniofacial development. The central thesis is that a combination of laser capture microdissection, and FACS, combined with microarrays and next generation sequencing can be used to efficiently achieve this goal. Microarrays, for example, with essentially complete gene representation can be used to rapidly determine the expression levels of every gene in laser capture microdissected elements of craniofacial development. A single experiment, therefore, provides a comprehensive analysis of the gene expression status of one component and a limited number of experiments examining each structure and cell type can create an atlas. A combination of structure, lectin staining, and transgenic GFP expression will be used to precisely identify specific compartments and lineages, including cells driving neural crest induction, nasal placodes and pits, lateral and medial facial eminences, neural crest and paraxial mesoderm cells, maxilla and mandibular processes, signaling centers, and the structures of palatogenesis. Next generation sequencing provides a digital readout of gene expression levels, cross-validates microarray data, gives additional valuable information concerning alternative processing, and can detect expression of genes not represented on arrays. We propose, therefore, to combine the use of microarray and nexgen sequencing in the creation of the craniofacial gene expression atlas. In addition, there are target amplification technologies that make possible the global analysis of gene expression in even single cells, which we propose to use to define molecular level heterogeneity present in the different cells of individual components. Using integrative bioinformatics analysis methodologies, gene expression modules reflective of variously differentiating components will be linked to gene networks implicated by shared structural, functional, and interactome features of the hundreds of genes already known to play individual roles in craniofacial development.

Genetic Tools and Resources for Orofacial Clefting Research

Orofacial clefting is one of the most common birth defects in humans, affecting approximately 1 in 700 live births. This frequency highlights the complexity of craniofacial morphogenesis, which requires precise regulation of gene expression changes, alterations in cell physiology and morphogenic movements. The mouse has played an instrumental role in advancing our understanding of the mechanisms that govern mid- face and palate development. Future progress, however, will require an increasingly sophisticated set of genetic models and tools. The overall goal of this project is to facilitate orofacial clefting research by generating new mouse genetic tools and by providing a repository of mouse strains critical for clefting research community. First, we will generate both inducible and constitutive Cre recombinase driver lines as new genetic tools. These new mouse strains will be designed specifically to support orofacial clefting research, with input from members of the Face Base consortium. These lines will be characterized in our established pipeline for evaluation and quality control of Cre-expressing mouse strains. Second, we will provide a repository for importation, cryopreservation, genetic quality control, and distribution of new and existing mouse models and tool strains important for orofacial clefting research. Such strains will include genetically engineered, spontaneously occurring, and ENU-induced models. New models for orofacial clefting will be actively solicited from members of the Face Base consortium and from the scientific community at large. We will create a public website for the Jackson Face Base Repository to promote the use of this resource and to facilitate access to mouse models and tools for orofacial research. Together, these projects will provide both Face Base members and the general scientific community with new research tools, novel genetic models, and comprehensive mouse repository services to enhance research in orofacial clefting.


Supplement Grant Number


Supplement Abstract
The objectives of this revision application are to identify and characterize craniofacial (CF) disorders in mice that are reliable genetic and physiological models for human craniofacial dysmorphologies and to share these models with the scientific public. Discovery of the genetic cause of CF disease in human populations is difficult due to extreme heterogeneity in humans and to the diversity of environmental and nutritional variables, all of which have a role in resulting CF diseases. Consequently, animal models with defined genetic backgrounds maintained in controlled environments are important for CF gene discovery. The goal of this revision application is to support CF research by providing new mouse models of CF dysmorphologies to FaceBase investigators, and the broader CF research community, via the FaceBase Mouse Repository. Utilizing our well-established program to screen for phenotypic deviants in The Jackson Laboratory’s large breeding and research colonies, our ongoing ENU mutagenesis program, and access to newly developed strains via our participation in the NIH-wide Knockout Mouse Phenotyping Program (KOMP2), we will identify, characterize phenotypically and genetically, and distribute novel mouse models of CF disorders. These will include the full range of CF phenotypes, including defects in skull morphology, dentition, vision, and hearing as well as models of orofacial clefting. For most models, we will identify the causative gene using high throughput sequencing technologies. Together, the models within this resource will provide critical tools for understanding the basis of CF disorders in humans.

Shape-Based Retrieval of 3D Craniofacial Data

Craniofacial malformations are among the most common structural malformations in humans. Researchers studying disorders of the craniofacial anatomy have many 3D imaging tools available to them, including computed tomography, magnetic resonance imaging, and 3D surface scans. Craniofacial researchers studying particular disorders are constructing large image databases of subjects in their studies. The Face Base Consortium will provide a central HUB for collection of data from numerous sites, enabling studies that would otherwise not be possible. Since shape is the critical factor in the classification of most craniofacial disorders, tools for analyzing 3D shape are essential to these studies. Quantitative shape descriptors allow for reproducible shape description, while similarity-based shape retrieval allows comparisons to be made between individuals or populations. The goal of this project is to develop tools for shape-based retrieval of 3D craniofacial image data. The specific aims of the project are to: 1) develop software tools that produce quantitative representations of craniofacial anatomy that can assist in the study of mid-face hypoplasia and cleft lip and palate; 2) develop tools for quantifying the similarity of craniofacial data between two individuals, between an individual and an average over a selected population, or between two populations; 3) develop mechanisms for organization and retrieval of multimodality 3D craniofacial data based on their quantitative representations; and 4) design and implement a prototype system for Craniofacial Information Retrieval (CIR) that incorporates quantification, organization, and retrieval; evaluate it on 3D craniofacial data and make it available to the Face Base HUB. The design of these tools and a pilot system will lead to a general methodology that is immediately applicable to studies of mid-face hypoplasia, cleft lip and cleft palate, but is also scalable and modifiable to all craniofacial abnormalities. NIDCR has awarded a supplement to this project to develop an Ontology of Craniofacial Development and Malformation (OCDM), and to demonstrate its usefulness for both the parent project and the FaceBase consortium as a whole. More information is available


The CranioGUI application provides web-based support for using our tools to analyze 3D facial mesh images.


Algum is an open-source web application developed at Carnegie Mellon University for generating Diamond filters using machine learning on marked up images.

Algum was originally developed by Shiva Kaul and was implemented using JRuby on Rails. It has been completely reimplemented by Jan Harkes using Python and the Django web framework.