Deep Neural Network For Prediction of Ageing

Exiting news for people working the field of ageing, biogerontology and computational chemistry. 

Ageing can be viewed in two ways

a) Ageing at physiological or organ system level
b) Ageing at cellular level

mostly ageing at cellular level is responsible for causing problems at organ system level. However, fighting ageing at cellular level by targeting single molecular pathway may not be sufficient enough due to complex dynamics in action that causes ageing to progress forward. Complexity of life can be unfolded by simplest molecules that causes an amplified effect in a bigger picture.

We know that cancer, Alzheimer's, cardiovascular problems and others are age related diseases that lead to problems. Pathways driven by proteins like m-TOR, sirtuins and others are been explored to fight ageing. Recently Insilico. Inc undertook a systematic studied on fundamental bio-molecules and their correlation with ageing. The paper can be downloaded from here.


The paper explains the development of a model based on 64,000 blood sample tests from human subjects. The model is based on 21 important blood test biomarkers that can be used for prediction ageing process in individuals. They have successfully identified five important biomarkers that play critical role in ageing.

i) Albumin
ii) Glucose
iii)Alkaline phosphatase
iv) Urea and
v)  Erythrocytes 

For common people they have availed a web portal called www.aging.ai where by providing the blood test results, the AI can predict the age of subject with surprising high accuracy. It clearly shows the advances in the computational biology towards developing deeper and complex correlations between disease and biomolecules. In Silico Inc. is currently making considerable effort in developing deep learning solutions in the field of regenerative medicine, embryonic development, cross species comparison and new drug discovery program.

Pharma.Ai is part of Insilico.Inc that is mainly responsible for use of deep neural networks in biomedical application. The group is currently working on two major projects of analyzing 

a) Correlation between transcriptomic data and ageing
b) Deep learned multi-disease marker and classifier 
c) Predictor of mechanism of action of multiple drugs based on multi-omics data 
d) Blood chemistry based marker for human ageing (One of the milestone was published recently as discussed above).

Recently Insilico Inc. is also planning to collaborate with other major institutions to find biomarkers for human ageing. Please check following recent press release

Insilico Medicine will provide advanced signaling pathway activation analysis services to evaluate differential changes between healthy tissues and those affected by disease as well as comparing tissues of different ages. In addition to signaling pathway analysis, parties intend to develop artificially-intelligent comprehensive biomarkers of human aging based on large sample data sets.

I-Robots for Age Diagnosis

There has been substantial boom in the field of artificial intelligence. It is important for people to understand that one can fight ageing with following perspectives:

a) Anti-ageing molecules/ Cosmetics

and also with

b) Artificial Intelligence (AI).

Fighting ageing with AI/ i-robot is thought to increase by tenfolds in next five years. This may help in fighting cancer and come up with newer and better algorithms on cancer diagnosis or mapping the efficacy of chemo on your smart phones. Mapping ageing with face scanning or taking selfies was the first step of Insilico Inc. As a next step, they have launched new website; where by using simple blood test an AI on the backend can predict the age of subjects. 

Please check here: It is a web portal that is scientifically tested and available for free usage. Moreover, it is developed with an aim of identifying markers that are responsible for ageing. People often discuss about food habits, drinking adequate amount of water, doing exercise and other activities for becoming healthy and fit. But does it really work? It certainly keeps human body fit and active but does it help in slowing the ageing process?

In order to answer such questions it is important to develop a correlation data between basic factors effecting ageing and dietary habits. However, to establish this relation only handful of information is available on the markers of ageing (for e.g. Telomeres, m-TOR, Sirtuins etc). Important thing to notice is that the "Blood" remains in contact with all the human tissues. This is an important material or matrix which carries common materials that are most essential to human cells and tissues with varied morphologies and activities. Hence, may be the answer to finding markers perhaps be hidden somewhere in this red fluid as shown by Insilico Inc. by launching algorithm that maps ageing based on blood test data.

In addition Aging.ai developed by Insilico Inc and RYNKL's  wrinkle reading app (as discussed previously); Microsoft has also developed a new age mapping web portal for free that answers a most basic question "How Old Do I look?"


Even though we ponder on the question "Is ageing a disease?". Well, there is no "Official" answer to this question yet. But the data shows that people are so obsessed with ageing that it took just 7 days to reach around 50 million users for microsoft's fancy tool (as discussed in video).  May be we don't need any official notification to testify that "Ageing is a disease".  

Leaders like Microsoft, Insilco Inc, SENSE foundation and other organisations are constantly finding newer ways to fight ageing by learning about how ageing happens.  Facial feature recognition AI by Microsoft and RYNKL may provide a fun time for people post their predicted age and post it on Facebook, Twitter or any other social media site. But apart from fun time this is concept of predicting ageing perhaps also  be applied to predicting the occurrence of age related disease like cancer.

Insilico Medicine launches Aging.AI -- deep-learned predictor of age trained on blood tests

Insilico Medicine launches Aging.AI -- deep-learned predictor of age trained on blood tests: Insilico Medicine launched aging.AI, a system allowing users to guess their age and gender by entering the results of their blood test.

Mobile App to Map Ageing

As you all must be knowing that researchers from Insilco Inc are constantly working to find newer ways to fight ageing. Researchers often study ageing at cellular and molecular level. The most obvious and prompt signature of ageing is "Face". People often use many skin care products such as powder, cream, lotions, face masks, anti-wrinkle creams etc. to prevent ageing.

It is astonishing that according to statista database global anti ageing market is increasing exponentially with time and it will be increasing to 154 million USD by 2021.


Figure 1: Skin Care Product Market

However, few important questions:

a) Does these products really work?
b) How should you quantitatively track the efficacy of the product you are applying?
c) Is it worth looking at your face for hours in mirror to find if the products work or not?

Interestingly, researchers have developed a new mobile app with an artificial intelligence that can quantitatively map ageing of your face. A team lead by Dr. Alex Zhavaronkov shows that it is possible to map ageing using this app. And see how the anti-ageing products are working over time.


The app is available for android users for free and can be downloaded on your android smart phones. For IOS users the app is going to be launched soon in February. Please register here for your notification of the app once it is launched. It was pretty amazing app so I tried it by myself and check my score to find out how does it works in figures shown below.


Although the pic shows a pretty glamorous girl by app..I am a guy. We need to upload the selfies as per the instructions given by Dr. Zhavaronkov and map ageing based on wrinkles. My score shows as 5.0 that indicates that my skin is healthy and I am beautiful (fewer wrinkle lesser number). This app can even help people to see if their anti ageing product is working or not.

Try this app it is fun and very useful

Please check this video for more details.


Moreover, Beauty.AI have updated their first list of winner based on age. The contest was judged by robot jury created by

a) RYNKL: Wrinkle Detector
b) MADIS: Model Alliance Digital Intelligence Scout
c) Symmetry: Master

Algorithm to Map Ageing

In addition to this app the Insilico Inc also launched a newly advanced algorithm to map ageing in humans by checking blood glucose and cholesterol levels. You can just feed in the data about your latest blood tests and AI will guess your age and sex. The AI is already trained for million samples.
The algorithm can guess sex, without using hormone levels, with 99% accuracy, and guess your age within a ten-year window at least 60% accuracy using ten key measurements found in a standard blood test

Dr. Zhavoronkov says“We want robots to look at human lifespan as a game,” he said — a game that artificial intelligence can help us win.

You can get more details about this news article here........ or at tech insider  or at eurekalert