In recent years, with the development of artificial intelligence technology and the deepening combination of medical industry and artificial intelligence technology, medical treatment has become one of the most important and fastest landing application scenarios of artificial intelligence technology. The application of AI in imaging assisted diagnosis, drug R & D and health management has become a clear development direction in the future.

As the most important sector in the pharmaceutical field, new drug R & D has shown a trend of slowing down in recent years due to problems such as high risk, high cost and long R & D cycle. However, the application of artificial intelligence technology makes it possible for the repetitive work that originally requires a lot of manpower in new drug R & D to be replaced by machines, thus shortening the R & D time, Reduced R & D costs.

Huizhong Research Institute summarizes and combs the global artificial intelligence new drug R & D enterprises in recent years, and comprehensively analyzes the overall situation of the industry in combination with relevant statistical data in recent years, in order to help the capital find good investment opportunities and jointly promote the healthy development of the medical industry.

This report includes the following contents:

1、 Industry background and market data

2、 Industrial Map

3、 Analysis of domestic related enterprises

4、 Analysis of overseas related enterprises

5、 Patent analysis

6、 Capital trend analysis

7、 Investment risk

1、 Industry background and market data

In July 2017, the State Council issued the development plan for a new generation of artificial intelligence, which specifically proposed to carry out large-scale genome recognition, proteomics, metabolomics and other research and new drug research and development based on artificial intelligence, so as to promote the intellectualization of medical supervision. The introduction of this policy has not only attracted more attention to the application of artificial intelligence in the field of new drug R & D, but also brought a new driving force to the development of the industry.

Summarize and sort out the global AI new drug R & D enterprises and comprehensively analyze the overall situation of the industry

Source: Deloitte

According to relevant reports released by Deloitte, the return on investment in new drug R & D in 2017 was only 3.2%, showing an obvious downward trend since 2010, while the cost of a new drug on the market was increasing year by year. By 2017, the cost of a new drug on the market had reached us $1.99 billion.

Summarize and sort out the global AI new drug R & D enterprises and comprehensively analyze the overall situation of the industry

Source: TDI ox.ac.uk

The R & D process of new drugs is complex and the R & D cycle is long. From the initial preliminary research and basic research to drug target research, biochemical research, pre clinical development, clinical trials (phase I, II and III), drug preparation stage and post marketing monitoring (phase IV), the average development cycle of a drug is more than 6 years, and the final success rate is less than 10%, However, artificial intelligence technology can quickly analyze compound structure, basic physiological mechanism and gene data, and process massive medical data, so as to greatly improve R & D efficiency, shorten R & D process, and save billions of dollars in R & D cost for the whole pharmaceutical industry.

Summarize and sort out the global AI new drug R & D enterprises and comprehensively analyze the overall situation of the industry

Source: tantalum data and online public data

From the perspective of artificial intelligence market, China has developed rapidly in recent years, from less than 10 billion yuan in 2016 to about 20 billion yuan in 2018, with a compound annual average growth rate of more than 40%. It is predicted that the total market scale of China’s artificial intelligence will exceed 40 billion yuan by 2020. At present, the application of artificial intelligence technology in the medical field in China is generally concentrated in the image assisted diagnosis sector, which still has a very large development space for the new drug R & D market.

However, in terms of more technological development, new drug discovery is still in the promotion stage of technology germination, the technology is relatively immature, and the maturity cycle of the technology may be relatively long in the future. There are some relatively mature enterprises and applications in the foreign market, while the overall domestic start is relatively late, and there is still a long way to go in the future.

2、 Industrial Map

According to the characteristics of artificial intelligence new drug R & D field, Huizhong Research Institute divides its structure into industrial chain and business territory, as shown in the figure below:

1. Industrial chain

Summarize and sort out the global AI new drug R & D enterprises and comprehensively analyze the overall situation of the industry

Source: tantalum data

The direction of AI + new drug R & D is relatively narrow, and the industrial chain contains few types of enterprises. The upstream mainly includes scientific research auxiliary platform, automation laboratory and big data and artificial intelligence tool platform. The midstream is the most mainstream drug development auxiliary platform, and the downstream is mainly drug R & D institutions. Some enterprises also independently develop drug development auxiliary platform while developing drugs. The flow relationship of the industrial chain is reflected in the flow from the bottom support technology or tools to R & D auxiliary tools, and then to R & D.

2. Business territory

2.1 classification according to research stage

From the research stage, artificial intelligence technology has gradually penetrated into all stages of new drug R & D, and has rich applications in the fields of basic research, drug target, biochemical activity, preclinical and clinical, and drug combination, among which the discovery of new targets, drug screening and optimization, and drug redirection are the most representative, At the same time, it is also the focus of artificial intelligence technology in the field of new drug research and development in various countries.

2.2 other applications of AI technology in new drug research and development

Source: tantalum data

In addition to the application in various stages of new drug R & D, there are other application scenarios related to drug R & D, such as automatic cell counting, traditional Chinese medicine identification, personalized medication, paper reading and information extraction, and automatic production related to automatic production, such as cell therapy, automatic production and automatic cell culture. All application scenarios have a certain support for new drug R & D.

2.3 disease direction and related technologies

AI is mainly used in new drug research and development in some serious and difficult to overcome diseases, such as cancer, neurodegenerative diseases, Alzheimer’s disease, autoimmune diseases, etc., but also has more applications in other high risk research and development related drugs, such as cardiovascular disease, Xie Ji’s disease, especially diabetes mellitus and bacterial infections.

From a technical point of view, in addition to the three core AI technologies of image recognition, natural language processing and machine learning, genomics data and cloud computing are also more important in the field of AI + new drug research and development.

3、 Analysis of domestic related enterprises

Huizhong Research Institute comprehensively collates domestic enterprises in relevant fields according to relevant data,

From the current situation of domestic enterprises, only 14 enterprises have layout in the field of artificial intelligence new drug R & D, and all of them are concentrated in Jiangsu, Zhejiang, Shanghai, Beijing and Guangdong.

On the whole, the most likely reason for the lack of entry of domestic enterprises is that the domestic development of this sector is relatively late and the technical entry barriers are high. Among these domestic enterprises, some enterprises (such as Taimei medical, medical data, Jiaxing Mairui medical, etc.) mainly focus on the medical information system business in the early stage, and gradually expand their business scope to the direction of artificial intelligence, This sector is not the most important and core business of the enterprise.

2. Industrial chain and business combing

According to the comprehensive sorting and analysis of the industrial chain and business map in this field, Huizhong research institute makes the following statistics on the business segments involved in domestic enterprises:

Summarize and sort out the global AI new drug R & D enterprises and comprehensively analyze the overall situation of the industry

Source: tantalum data

In addition to baiaozhi as a research assistance platform, other 13 companies provide drug development assistance business, in which medical think tank provides drug development assistance and research assistance platform services at the same time.

Summarize and sort out the global AI new drug R & D enterprises and comprehensively analyze the overall situation of the industry

Source: tantalum data

According to the analysis of the business segments involved in the enterprise, at present, the layout of domestic enterprises is relatively concentrated in innovation target identification and compound screening, and at least 4 of the 14 enterprises have layout. Paper reading and information extraction, selection of leading compounds, drug molecular involvement, pharmacovigilance Drug synthesis route design and other businesses are distributed in at least 2 of the 14 enterprises, while other businesses are relatively few.

4、 Analysis of overseas related enterprises

Huizhong Research Institute sorted out AI + new drug R & D related enterprises in other countries around the world according to relevant data. According to incomplete statistics, there are a total of 149 (including some M & A enterprises, excluding the relevant new drug R & D platforms independently developed by giant pharmaceutical companies. For a detailed list, see Appendix 1).

1. Statistics on the number and distribution of global enterprises

Summarize and sort out the global AI new drug R & D enterprises and comprehensively analyze the overall situation of the industry

Source: tantalum data

According to the current data, the United States has the deepest layout in AI + new drug R & D, with 86 enterprises, accounting for more than 50%, followed by the United Kingdom and Canada, with 27 and 11 enterprises respectively. It is also distributed in France, Germany, Japan, Canada and other places.

2. Statistics of enterprise establishment time

Source: tantalum data

In terms of the time of establishment, except that some old enterprises before 2000 took AI + new drug R & D as the direction of transformation or business expansion, the vast majority of enterprises were registered after 2000, especially after 2011, the number of enterprises registered each year exceeded 10 (the statistical data in 2018 is not complete for reference), and from 2008 to 2016, The number of enterprises established each year shows an obvious gradual upward trend, and began to decline in 2017.

From the layout of AI + new drug R & D direction of overseas companies, it is obviously different from domestic enterprises. Overseas enterprises focus more on clinical trial patient group screening, personalized medication, drug redirection, paper reading and information extraction, especially the stratification technology of clinical trial patient subtypes and groups.

5、 Patent analysis

Huizhong research institute makes statistics on all patent applications of 163 companies at home and abroad in the field of AI new drug R & D, and summarizes the patents in relevant fields as follows:

According to the statistics of patent applications of 163 companies at home and abroad in the field of AI new drug R & D, the number of patents in this field has increased significantly since 2011, peaked in 2016 and decreased slightly in 2017.

It can be seen that in recent years, the enthusiasm for R & D in this field has maintained an upward trend as a whole. In 2016, the enthusiasm for R & D and the intensity of technical competition reached the peak, and the competition slowed down slightly in 2017 (due to the particularity of patent data, the statistical data in 2018 is not perfect, and this part is for reference only).

According to the statistics on the growth rate of patent applications of Companies in the field of AI new drug R & D, the field of AI new drug R & D has maintained a high growth rate as a whole, and the growth rate has slowed down after 2013.

The main reason for this phenomenon is the gradual increase of the patent base of the industry itself. At the same time, for new entrants, the entry difficulty is gradually increasing.

According to the distribution of patents involved in the field, organic pharmaceutical formula patents occupy the most important part. Therefore, it can be seen that the main service object of AI new drug research and development is chemical drugs.

The field with the second largest proportion is the special-purpose data calculation or processing method patents, which is the most competitive technology field in the field of AI pharmaceutical combination technology and the most important field for enterprises to realize their own competitive advantage.

Other representative fields include identification methods of enzymes or microorganisms, pharmaceutical preparations with unusual active ingredients (such as those with various biological factors as active ingredients), computer systems for biological models, automatic analysis technology of unconventional materials, physical image and character recognition technology patents, etc.

After ranking the number of patent applications of Companies in the field of AI new drug research and development, it can be seen that the company with the largest number of patent applications in this field is pharnext (French biopharmaceutical company, the main treatment field is nervous system diseases). The number of patent applications of the company is twice that of the second ranked enterprise, and the company has a strong technical competitive advantage compared with other companies in the field.

Berkeley lights, a biotechnology company in the United States, and Charles River, a preclinical outsourcing company in the United Kingdom, ranked second and third, with 212 and 181 patent applications respectively. Other companies with more than 100 patent applications are ubiome, a microbial gene sequencing company in the United States, and e-therapeutics, an innovative drug R & D company in the United Kingdom.

According to the statistics of the top 3 companies in the number of patent applications, the number of patent applications of Berkeley lights has increased significantly in the past five years, and it is expected to have strong technology explosion potential in the future. The number of patent applications of pharnext and Charles River decreased significantly this year, and the innovation vitality of the two companies in recent years was insufficient.

6、 Capital trend analysis

Huizhong research institute makes statistical analysis on the financing history of 163 domestic and foreign companies in the field of AI new drug R & D, and the relevant information is summarized as follows:

Source: tantalum data, CrunchBase, cbinsight

As of March 2019, more than 140 AI + new drug R & D companies at home and abroad had obtained financing, with a cumulative number of 342 transactions and a total financing of US $48.3 billion. Among them, the cumulative financing amount of foreign enterprises is 45.5 billion US dollars, and only one domestic enterprise has not had financing, with a total cumulative amount of 2.792 billion US dollars. More than 90% of enterprises in the field have financing history, and the capital heat is very high.

From the distribution of events invested every year, from 2001 to 2018, the number of transactions and financing amount increased year by year. Among them, the number of transactions in 2017 was the same as that in 2018, but the financing amount in 2018 was the highest, reaching US $1.8 billion, and the number of invested enterprises accounted for more than 40%.

From the perspective of investment rounds, most of the invested projects are in the seed round, the number of transactions accounts for more than 70%, and the financing amount accounts for nearly 30%. Large amount financing is mainly concentrated in round B, round C and round D. In addition, countries attach great importance to the development of this field. The number of investment through government subsidies is as high as 32 times, and the financing amount exceeds US $600 million.

From the public historical data, there are 445 investment institutions involved in AI + new drug R & D, including 11 with more than 5 transactions, and ame cloud ventures, Khosla ventures and Google ventures have the most transactions. It is worth mentioning that Bill Gates personally participated in five trading events, all of which were invested in nimbus therapeutics.

7、 Investment risk

1. AI + lack of new drug R & D talents

According to relevant surveys, 41% of the 330 drug R & D scientists do not understand AI technology, so they can not use AI to screen new drugs. The lack of AI talents and the lack of understanding of AI technology by R & D personnel may lead to the integration between disciplines can not reach a good level in just a few years. The lack of interdisciplinary talent training in the future may restrict the development of the field.

2. Complex biology makes drug development more difficult than expected

Due to the extremely complex characteristics of biology itself, new molecules that can work in theory and models may have various unpredictable results in the human body and may have complex reactions with other molecules. At the same time, individual differences further increase the complexity of drug research and development, which also increases the difficulty of the application of AI technology.

3. Insufficient quality data

The number of new drugs approved for use on the market is limited, and the amount of these data is far from enough. The new drug R & D rules are not clear, the data is not clear and full of high uncertainty, which brings great obstacles to the artificial intelligence based on data set. At the level of patent protection, drug patents, especially the core compound patents, pharmaceutical companies usually do not disclose too much data, let alone share the most popular target data they are currently developing, which makes the amount of data available less than theoretically, and further limits the availability of high-quality data.

Leave a Reply

Your email address will not be published. Required fields are marked *