DatAnon

DatAnon is very pleased to have been invited to the CancerHacker Lab Accelerator. Together with so many talented startup founders, we are excited to help tackle the urgent problem of improving cancer diagnostics and treatment for all patients. Our first platform addressing these pressing needs will be launched soon for user feedback - watch this space in the coming weeks!

DatAnon's methods detect responders in #clinicaltrials, particularly in #cancer. The graphic simplifies the main goal - in a clinical trial, all patients in the treatment arm get the drug. Some respond positively, and some do not. Identifying which patients respond can dramatically improve the chances of passing more compounds, which are more efficacious in responders. This is exactly the purpose of our technology. It has the potential to get more drugs approved with improved treatment effects. Which has the ultimate goal of improving the lives of cancer patients. We'll be posting our pre-print and demo software very soon! #precisionmedicine #personalizedmedicine #oncology #patientresponder

We're launching #patientresponder soon to give insight into drug response by cancer type: Using cell line tumor treatment data generated by the Dependency Map project, DatAnon uses our technology to create gene expression signatures and subgroups based only on the geometry of expression. For over 4500 drugs and nearly 500 cell lines. These show responders by testing efficacy via leave one out prediction accuracy, with permutation testing for creating p-values of association. There's a lot under the hood to explain. Use cases: 1) Find potential subgroups of response for a drug in a given cancer type ➡ Query multiple drugs in a given cancer type to find efficacious compounds 2) Find responders across multiple cancer types for a single drug ➡ Or across all drugs 3) Much more in the months to follow Free to use for researchers and academics. #cancer #oncology #datanonymization #precisionmedicine #precisiononcology

DatAnon is the Data Anonymization company. There is a overwhelming need to share and analyze data anonymously. What reasons do you think are the most important for data anonymization? 1) Protect IP - companies and individuals cannot share data that constitutes intellectual property, or reveals any part of their valuable data assets. 2) Preserve privacy - individuals in clinical trials, etc. must be kept private. Thus, any identifying data, which may include the data itself, cannot be revealed to other parties. 3) Comply with regulations - Data sharing of personal biomedical information across borders is forbidden for many countries, even if the data cannot identify individuals. GDPR, HIPAA, etc. have additional requirements. 4) Protect against data breaches - There have been numerous instances of data breaches from mobile phone operators, banks, even one of the credit bureaus! When data is anonymized, a breach becomes impossible. Did I miss anything important? Let me know in the comments!! #anonymization #clinicalresearch #precisionmedicine #clinicaltrials #dataanalytics #data #pharmaceuticals #pharma #drugdiscovery

What is the difference between data de-identification and anonymization?? Typically, de-identification means that an individual cannot be re-identified in the dataset. So "John Doe" or a social security number might be replaced with a hash function. Be careful though, the data itself might be able to re-identify the individual! Like genomics... Anonymization usually means the data itself is changed in a way that the original values cannot be recovered. Data masking, differential privacy and synthetic data are common approaches. Encryption makes data unreadable, but if the data can be decrypted you must be cautious! 🔁 At DatAnon , we use one-way geometrical transformations, which cannot be reverse engineered --> the original data is not recoverable! And due to embedding important geometrical attributes, the transformations give information for our ML layer to find subgroups of patient responders. #anonymization #clinicalresearch #cancer #precisionmedicine #clinicaltrials #dataanalytics #data