NeuroDevNet was a federal Network of Centres of Excellence with one big bet: that researchers studying autism, cerebral palsy, fetal alcohol spectrum disorder, and the genetic and imaging methods underlying all three would publish more — and better — if they actually published together.
This is what happened to that network over four years. 316 researchers, 88 publications, 2,903 collaborations — every paper re-classified by an LLM from its abstract so we can see what each cluster was actually working on.
The first NeuroDevNet co-publication appeared on April 29, 2010 in Human Brain Mapping — a methods paper that used MRI to study cortical development. Four authors. The first edges of the network.
By year-end, just 13 unique researchers had published anything under the network's umbrella. Most sat at the centre of well-established groups. The network in its first year was barely a network at all — a handful of clusters that hadn't yet found each other.
End of 2010: 13 researchers · 1 publication
9 were direct NeuroDevNet members; 4 were external co-authors. The first paper was firmly in the Genetics & Imaging theme — methodological infrastructure that would pay off later.
Then the network exploded. By end of 2011, 122 researchers had co-published. By 2012, 166. Each newcomer brought their own collaborators; each publication created new edges.
The thematic clusters hardened. The 2011 papers were dominated by a pair of genetics findings that became the network's two most-cited works: an autism CNV study in Nature (later cited 2,000+ times) and the PTCHD1 locus discovery in Science Translational Medicine. Cerebral palsy researchers were publishing in parallel about gross motor function and motivation in adolescents with CP. FASD, meanwhile, sat alone — its papers connected to no one outside the fetal-alcohol group.
End of 2012: 166researchers 38publications
The two highest-impact 2011 papers were classified as ASD-Genetics by abstract analysis: rare copy-number variants and PTCHD1 mutations in autism & intellectual disability.
The next two years were the network's adolescence. 102 new researchers in 2013, another 104 in 2014. The clusters bound together — and re-classified by abstract, the picture is sharper than the original journal-based view suggested.
After cross-referencing the NeuroDevNet authors' own publication list against PubMed and OpenAlex, 81 papers are classified — and within each big theme, distinct sub-areas emerge.
Autism research (24 papers) splits cleanly: nine genetics papers (CNVs, PTCHD1, ASTN2 — the high-citation core), five baby-sibling early-ID studies, four clinical/family papers (parenting stress, advocacy), three sensory-processing papers, two phenotyping/heterogeneity papers, and one neuroimaging review.
Cerebral palsy (21 papers) is intervention-heavy: eight intervention trials (stem cells, computer-play motor therapy, brain stimulation, hyperbaric oxygen), four quality-of-life papers (leisure, mastery, family perspectives), four motor-function papers (GMFCS/MACS), three etiology papers (chorioamnionitis, perinatal stroke), and two preterm-imaging papers.
Knowledge Translation (12 papers) is half ethics: five neuroethics papers (biomarker screening, incidental findings, transition ethics), four KT practice papers (knowledge brokers, mobilisation models), two community-engagement papers (Indigenous health), one care-transition paper.
Add FASD (7 papers — prenatal-alcohol-exposure biology, eye-movement deficits), Genetics & Imaging methods (6 — including the ENIGMA consortium), and a small Intervention bucket (2 — therapeutic hypothermia, antipsychotic prescribing).
FASD finally connected to the rest. A 2013 paper on HPA/dopamine alterations after prenatal alcohol exposure linked the green cluster to ASD-genetics work; a 2013 review of FASD eye-movement deficits brought in cerebral-palsy collaborators. The isolated enclave became part of the fabric.
A pink Knowledge Translation cluster also emerged in the bottom-right — researchers studying not the disorders themselves but how to move evidence from research into care: youth transitions, family decision-making, neuroethics.
By May 2014: 316researchers 88pubs 2,903collaborations
The 87-author Hum Mol Genet paper has been excluded from this view — it was a single-paper supercluster that visually overwhelmed the rest.
Most researchers stayed inside their cluster. But a small set co-published across multiple themes — and they are the reason this became a network rather than a federation of subgroups.
At the centre sat Wendy Roberts — the only researcher whose authored papers spanned FASD, cerebral palsy, AND autism research simultaneously. Each of those papers shows up in a different colour cluster, and her node is the deepest blended hue in the visual.
Stephen Scherer (autism genetics) co-authored ASD-CNV and PTCHD1 papers that connected the autism-research and Genetics & Imaging clusters. Michael Shevell anchored the cerebral-palsy cluster, with the densest set of inter-theme collaborations between CP and ASD running through his joint papers.
Sterling Clarren and Daniel Goldowitz anchored the green FASD cluster with the prenatal-alcohol-exposure work — animal models, HPA/dopamine biology, gene-environment. Their trainees (including Carmen Rasmussen) appear as a green-purple gradient on the boundary between FASD and CP, where joint work on motor and cognitive outcomes lives.
David Phipps seeded the pink Knowledge Translation cluster late but fast — nine of his publications classified as KT in 2013–2014, on community engagement, evidence-based health-services research, and ethics in care for youth with neurodevelopmental conditions.
These researchers' nodes use blended hues — combining the colours of every theme their papers touched. A node bridging ASD-blue and CP-purple takes on a violet tint. The deeper the blend, the more cross-domain the researcher.
Six themes, each with its own colour. Autism Research remained the largest and most central. Cerebral Palsy grew significantly — by 2014 it had radiated into the ASD core through shared work on pediatric neuroimaging. FASD, once isolated, had become integrated. Knowledge Translation appeared from nothing as a coherent fifth theme. Genetics & Imaging fed methods into all of them.
The studied populations show the same story: Children with ASD (17 papers) and Children with CP (17) tied for primary focus, followed by Adolescents (11), Caregivers/Families (10), and Infants/Neonates (7). The two most-cited papers — the ASD-CNV Nature paper (2,044 citations) and the ENIGMA imaging consortium (867) — were both genetics-and-imaging methods works whose findings flowed back into the disorder-specific clusters.
The hypothesis had its answer: yes, organising researchers as a network changes their publication patterns. Not all of them, but enough to matter.
Every node carries the researcher's last name — this is public authorship data; the relationships are real, and so are the people who built them.
Drag the timeline to scrub through the network's growth. Hover any node to see who they collaborated with — and which themes their actual papers fell into. Click a node to lock the view. Drag the canvas to pan, scroll to zoom.
Switch tabs to see the data through other angles. Themes & Populations shows a Sankey flow from theme → sub-theme → studied population — what each cluster is actually researching, in whom. Timeline stacks publications by quarter, coloured by LLM-classified theme — press Play to watch the chart draw across time.
Each researcher's tooltip shows their theme distribution — counted from the actual papers they co-authored, not just the journals they published in. A researcher with 9 ASD papers and 1 KT paper will read as primarily blue with rose flecks.
The visualisation is now fully interactive. There's no more scrolling story; this is yours.