Portrait of Julie Hussin

Julie Hussin

Associate Academic Member
Associate Professor, Université de Montréal
Research Topics
Computational Biology
Data Mining
Deep Learning
Medical Machine Learning
Multimodal Learning
Website
Google Scholar
LinkedIn
X

Biography

Julie Hussin is an associate professor in the Faculty of Medicine at Université de Montréal (UdeM) and a researcher at the Montréal Heart Institute. She is a Junior 2 Research Scholar funded by the Fonds de Recherche du Québec en Santé (FRQS) and chair of the graduate program in bioinformatics at UdeM.

Trained in statistical and evolutionary genomics, Hussin has significant experience in handling multi-omics datasets from large population cohorts. Her work in computational biology is relevant to medical and population genomics, fields in which she has contributed to several methodological advances. Her interdisciplinary work, which aims to develop innovative tools for precision medicine, focuses on improving risk prediction and the management of cardiometabolic disease, particularly heart failure.

Her approaches integrate various data types, such as clinical, genetic, transcriptomic, proteomic and metabolomic data, to uncover new insights into the biological determinants of heart disease, notably through unsupervised learning techniques. In the context of the COVID-19 pandemic, Hussin’s group also led the development of data science algorithms to analyze viral genetic data, aid viral surveillance efforts, and study host-pathogen interactions and viral evolution.

Her work also focuses on the interpretability, generalizability and fairness of machine learning algorithms in health research. Hussin is dedicated to promoting fair, safe and transparent AI in health research, striving for inclusivity and representation to ensure her work benefits all segments of the population. Her expertise also extends to the field of fair, safe and transparent AI for health research. She teaches several undergraduate and graduate courses in computational biology and population genetics, as well as machine learning for genomics. Prior to joining UdeM as a professor, she was a Human Frontier Postdoctoral Fellow at the Wellcome Trust Centre for Human Genetics at the University of Oxford (Linacre College), and a visiting fellow at McGill University.

Current Students

Cantin Baron
PhD - Université de Montréal
Website
Google Scholar
David Hamelin
PhD - Université de Montréal
Fatima Mostefai
PhD - Université de Montréal
Alexis Nolin-Lapalme
PhD - Université de Montréal
Simon Paquette
Master's Research - Université de Montréal
Github
Camille Rochefort-Boulanger
PhD - Université de Montréal
Co-supervisor :
Yoshua Bengio
Github
Matthew Scicluna
PhD - Université de Montréal
Co-supervisor :
Guy Wolf
Github
Google Scholar

Publications

Diet Networks: Thin Parameters for Fat Genomics
Adriana Romero Soriano
pierre luc carrier
Akram Erraqabi
Tristan Sylvain
Alex Auvolat
Etienne Dejoie
Marc-André Legault
Marie-Pierre Dubé
Julie Hussin
Yoshua Bengio
Learning tasks such as those involving genomic data often poses a serious challenge: the number of input features can be orders of magnitude… (see more) larger than the number of training examples, making it difficult to avoid overfitting, even when using the known regularization techniques. We focus here on tasks in which the input is a description of the genetic variation specific to a patient, the single nucleotide polymorphisms (SNPs), yielding millions of ternary inputs. Improving the ability of deep learning to handle such datasets could have an important impact in medical research, more specifically in precision medicine, where high-dimensional data regarding a particular patient is used to make predictions of interest. Even though the amount of data for such tasks is increasing, this mismatch between the number of examples and the number of inputs remains a concern. Naive implementations of classifier neural networks involve a huge number of free parameters in their first layer (number of input features times number of hidden units): each input feature is associated with as many parameters as there are hidden units. We propose a novel neural network parametrization which considerably reduces the number of free parameters. It is based on the idea that we can first learn or provide a distributed representation for each input feature (e.g. for each position in the genome where variations are observed in data), and then learn (with another neural network called the parameter prediction network) how to map a feature's distributed representation (based on the feature's identity not its value) to the vector of parameters specific to that feature in the classifier neural network (the weights which link the value of the feature to each of the hidden units). This approach views the problem of producing the parameters associated with each feature as a multi-task learning problem. We show experimentally on a population stratification task of interest to medical studies that the proposed approach can significantly reduce both the number of parameters and the error rate of the classifier.
2017-01-01
ICLR.cc/2017/conference (poster)
openreview.net
openreview.net
Diet Networks: Thin Parameters for Fat Genomics
Adriana Romero Soriano
pierre luc carrier
Akram Erraqabi
Tristan Sylvain
Alex Auvolat
Etienne Dejoie
Marc-André Legault
Marie-Pierre Dubé
Julie Hussin
Yoshua Bengio
Learning tasks such as those involving genomic data often poses a serious challenge: the number of input features can be orders of magnitude… (see more) larger than the number of training examples, making it difficult to avoid overfitting, even when using the known regularization techniques. We focus here on tasks in which the input is a description of the genetic variation specific to a patient, the single nucleotide polymorphisms (SNPs), yielding millions of ternary inputs. Improving the ability of deep learning to handle such datasets could have an important impact in medical research, more specifically in precision medicine, where high-dimensional data regarding a particular patient is used to make predictions of interest. Even though the amount of data for such tasks is increasing, this mismatch between the number of examples and the number of inputs remains a concern. Naive implementations of classifier neural networks involve a huge number of free parameters in their first layer (number of input features times number of hidden units): each input feature is associated with as many parameters as there are hidden units. We propose a novel neural network parametrization which considerably reduces the number of free parameters. It is based on the idea that we can first learn or provide a distributed representation for each input feature (e.g. for each position in the genome where variations are observed in data), and then learn (with another neural network called the parameter prediction network) how to map a feature's distributed representation (based on the feature's identity not its value) to the vector of parameters specific to that feature in the classifier neural network (the weights which link the value of the feature to each of the hidden units). This approach views the problem of producing the parameters associated with each feature as a multi-task learning problem. We show experimentally on a population stratification task of interest to medical studies that the proposed approach can significantly reduce both the number of parameters and the error rate of the classifier.
2017-01-01
ICLR.cc/2017/conference (poster)
openreview.net
openreview.net
Diet Networks: Thin Parameters for Fat Genomic
Adriana Romero Soriano
pierre luc carrier
Akram Erraqabi
Tristan Sylvain
Alex Auvolat
Etienne Dejoie
Marc-André Legault
M. Dubé
Julie Hussin
Yoshua Bengio
Learning tasks such as those involving genomic data often poses a serious challenge: the number of input features can be orders of magnitude… (see more) larger than the number of training examples, making it difficult to avoid overfitting, even when using the known regularization techniques. We focus here on tasks in which the input is a description of the genetic variation specific to a patient, the single nucleotide polymorphisms (SNPs), yielding millions of ternary inputs. Improving the ability of deep learning to handle such datasets could have an important impact in precision medicine, where high-dimensional data regarding a particular patient is used to make predictions of interest. Even though the amount of data for such tasks is increasing, this mismatch between the number of examples and the number of inputs remains a concern. Naive implementations of classifier neural networks involve a huge number of free parameters in their first layer: each input feature is associated with as many parameters as there are hidden units. We propose a novel neural network parametrization which considerably reduces the number of free parameters. It is based on the idea that we can first learn or provide a distributed representation for each input feature (e.g. for each position in the genome where variations are observed), and then learn (with another neural network called the parameter prediction network) how to map a feature's distributed representation to the vector of parameters specific to that feature in the classifier neural network (the weights which link the value of the feature to each of the hidden units). We show experimentally on a population stratification task of interest to medical studies that the proposed approach can significantly reduce both the number of parameters and the error rate of the classifier.
2016-11-04
ArXiv (preprint)
arxiv.org
arxiv.org