Insight Stats
Insight Stats is a Sydney based biotechnology and bioinformatics focused consultancy. Statistical analysis, ML and data mining minus the hype and buzzwords, stemming from hands-on experience with hundreds of studies. Straddling both industry and research including in a quality accredited environment, we have a broad network of collaborators including analytical chemists, mass spectrometry specialists, data mining and AI, biotech and clinical professionals.
Helping you make key decisions based on your complex datasets.
About
Dana Pascovici has a PhD in Mathematics from MIT, and a bachelor degree in Mathematics and Computer Science from Dartmouth College, NH. She spent 12 years as the biostatistician at the Australian Proteome Analysis Facility, and had a recent 3-year stint as research scientist/biostatistician at the CSIRO. In 2020 she founded Insight Stats (ABN 40 954 430 786), consulting for several biotechnology start-ups and research collaborations.
Case studies
Breast cancer lipid signatures
Working with ASX-listed diagnostics company on developing their lipid breast cancer blood test using mass spectrometry data over consecutive cohorts, including supporting the company IP and milestone publications.
Stability and analytical considerations
Consulting for startup driving innovations in blood diagnostics beyond plasma, with a focus on personalised medicine and remote testing. Supporting the company analytical/stability studies and VAMS IP portfolio publications.
Impactful research in neurodegeneration
Collaboration with researchers in neurodegeneration from Sydney University. Supporting omics discovery and creating bespoke interactive web tools for TDP-43 research showcased in key publication (Nature Comms. 2024).
Multiple sclerosis diagnostics development
Ongoing collaboration with start-up Proheme Diagnostics on marker selection, analytical and discovery workflow, diagnostics development.
Areas of expertise
Experiment design and general statistics
Hands on expertise ranging from the simple through to the complex. Applied to small clinical trials data, drug activity and potency, practical classification and numerous research experiments.
Biomarker discovery
Concrete experience with ML workflows including variable selection in complex biological datasets (RNA, protein, metabolites and lipids).
Analytical considerations
ML workflows are fine, but to translate your features must meet analytical requirements and considerations of stability and reproducibility. Data mining with an eye to translation to the clinic.
Data harmonisation and privacy
The bane of every analyst and start-up alike, your crucial data will come in infinitely messy forms. Providing advice on data harmonisation for analysis/sharing/validation, or synthetic/mock data generation for privacy considerations and application development.
Diagnostic algorithm design and optimisation
Practical and research experience in the area of developing panel classifiers for better diagnostics or prognostics, for immunoassays, mass spectrometry or multi-omics data.
Quality and IP considerations
Experience with generating software workflows in a NATA ISO 17025 certified environment. Practical experience and advice on IP considerations.
Contact
Reach out via referral or LinkedIn.