World Data Analysis

Publicly available protein databases: The game-changer medical infrastructure /!\ This is likely to become one of the most important medical and biotech infrastructure layers of the decade. While most attention goes to chips, GPUs, and frontier models, the real long-term AI infrastructure in medicine may actually be biological data itself. Public protein databases are becoming the digital backbone of drug discovery, diagnostics, protein engineering, and precision medicine. Interpretation of the chart The most explosive signal is AlphaFold DB. It jumps from roughly: ~1M structures in 2023 to ~200M in 2025. That is extraordinary. Meanwhile: UniProt remains above 100M and continues rising PDB grows steadily from ~140K to ~200K They are built through a broad ecosystem of: academic laboratories hospitals and clinical centers structural biology consortia pharmaceutical R&D cryo-EM and X-ray crystallography facilities AI-based structure prediction systems In other words, this is scientific infrastructure compounding in public view. What does this actually mean? A protein database stores: amino acid sequences experimentally resolved 3D structures AI-predicted folds binding pockets mutations disease associations enzyme functions The three pillars in the chart represent different layers of this stack: UniProt: the core sequence and functional annotation layer PDB: experimentally validated 3D protein structures AlphaFold DB: AI-predicted 3D structures at planetary scale Think of it as: the Bloomberg terminal of biology Who contributes to it? This infrastructure is built by a global public-private knowledge network. 1) Universities and research labs: MIT, Stanford, Oxford, EMBL, Max Planck, and thousands of labs globally. 2) Public research institutions: NIH, EMBL-EBI, and national genome programs. 3) Hospitals and clinical research centers: Especially for disease-linked mutations and rare protein variants. 4) AI labs: DeepMind, Isomorphic Labs, and computational biology teams. 5) Pharma and biotech: For target validation, protein-ligand mapping, antibodies, and therapeutic proteins. Where do these proteins come from? They emerge from: genome sequencing cancer biopsies pathogen surveillance antibody discovery rare disease studies plant biology synthetic biology microbial screening Every new pathogen, mutation, or disease mechanism can rapidly expand this public infrastructure. Why is this such a big deal? Because it compresses the cost and time of biological discovery. 1) Faster drug discovery: Known targets make molecule design dramatically faster. 2) Better cancer diagnostics: Protein mutations can be linked to treatment pathways. 3) Precision medicine: Patient-specific variants can be interpreted faster. 4) New industrial enzymes: For food, agriculture, materials, and climate tech. 5) Protein engineering: Antibodies, enzymes, and therapeutic proteins can now be designed at scale. What does this lead to next? The downstream effects are enormous: AI-designed drugs biologics acceleration rare disease therapeutics climate biotech enzymes protein-based materials synthetic biology platforms faster vaccine design Source: @Stanford

@CindySigal50395 @MoneyQuotesX and for @MoneyQuotesX too

@CindySigal50395 @MoneyQuotesX Thanks, I hope we do it someday, I’ll buy you a coffee.

Client: “I want to lose weight" Consultant: “Done” The rise of GLP-1 drugs is turning obesity treatment from a long-term lifestyle struggle into something that increasingly looks like a pharmaceutical solution The first chart shows something very important about the economics of GLP-1 and obesity drugs in the United States. Even after the huge Ozempic and Wegovy boom, most U.S. states still do not fully cover GLP-1 obesity treatments under Medicaid programs. A large part of the country either: * does not cover these drugs at all or * only covers older weight-loss medications. This matters because GLP-1 demand may still be in a relatively early phase. If broader insurance coverage expands in the future, the market could become much larger than it already is. At the same time, this also shows how difficult the economic side of healthcare can be. These drugs are medically promising, but they are also extremely expensive for public health systems. Source: @pewresearch by Mia Hennen

@InvestMetaAI Thanks for the input

@World_Data_A BYD and CATL scale faster because the logistics layer already exists. Jebel Ali transshipment, DP World's India terminals, bonded transit corridors. That's what enables the velocity these charts show.

. Chinese companies are becoming increasingly dependent on overseas markets The charts show how overseas revenues are growing much faster for several major China-based firms, especially after 2021. A few things stand out: BYD’s overseas revenue expanded rapidly in only a few years, while total revenue also surged. The company is no longer relying mainly on the domestic Chinese market. Geely also shows a visible increase in international exposure, though domestic sales still remain dominant. In the lower chart, firms such as Luxshare, BYD, Midea, CATL, and SAIC Motor all recorded large jumps in overseas revenues between 2021 and 2024. This matters because it shows Chinese industrial expansion is no longer only export-volume driven. Companies are increasingly building: foreign sales networks, manufacturing capacity abroad, logistics systems, after-sales ecosystems, and localized production chains. Another important point is diversification. These firms are spread across: EVs, batteries, electronics, appliances, industrial equipment, and chemicals. That suggests the globalization process is broadening beyond a single sector. In many ways, Chinese firms now resemble the earlier international expansion phases of Japanese and Korean conglomerates, but the scaling speed appears significantly faster, partly because they can leverage existing global supply chains, digital commerce, and China’s very large domestic manufacturing base simultaneously. Source: @rhodium_group by @CBoullenois, Malcolm Black and Alessia Caruso

@snowjacuzzi 👍

@World_Data_A 🤔 interesting

Publicly available protein databases: The game-changer medical infrastructure /!\ This is likely to become one of the most important medical and biotech infrastructure layers of the decade. While most attention goes to chips, GPUs, and frontier models, the real long-term AI infrastructure in medicine may actually be biological data itself. Public protein databases are becoming the digital backbone of drug discovery, diagnostics, protein engineering, and precision medicine. Interpretation of the chart The most explosive signal is AlphaFold DB. It jumps from roughly: ~1M structures in 2023 to ~200M in 2025. That is extraordinary. Meanwhile: UniProt remains above 100M and continues rising PDB grows steadily from ~140K to ~200K They are built through a broad ecosystem of: academic laboratories hospitals and clinical centers structural biology consortia pharmaceutical R&D cryo-EM and X-ray crystallography facilities AI-based structure prediction systems In other words, this is scientific infrastructure compounding in public view. What does this actually mean? A protein database stores: amino acid sequences experimentally resolved 3D structures AI-predicted folds binding pockets mutations disease associations enzyme functions The three pillars in the chart represent different layers of this stack: UniProt: the core sequence and functional annotation layer PDB: experimentally validated 3D protein structures AlphaFold DB: AI-predicted 3D structures at planetary scale Think of it as: the Bloomberg terminal of biology Who contributes to it? This infrastructure is built by a global public-private knowledge network. 1) Universities and research labs: MIT, Stanford, Oxford, EMBL, Max Planck, and thousands of labs globally. 2) Public research institutions: NIH, EMBL-EBI, and national genome programs. 3) Hospitals and clinical research centers: Especially for disease-linked mutations and rare protein variants. 4) AI labs: DeepMind, Isomorphic Labs, and computational biology teams. 5) Pharma and biotech: For target validation, protein-ligand mapping, antibodies, and therapeutic proteins. Where do these proteins come from? They emerge from: genome sequencing cancer biopsies pathogen surveillance antibody discovery rare disease studies plant biology synthetic biology microbial screening Every new pathogen, mutation, or disease mechanism can rapidly expand this public infrastructure. Why is this such a big deal? Because it compresses the cost and time of biological discovery. 1) Faster drug discovery: Known targets make molecule design dramatically faster. 2) Better cancer diagnostics: Protein mutations can be linked to treatment pathways. 3) Precision medicine: Patient-specific variants can be interpreted faster. 4) New industrial enzymes: For food, agriculture, materials, and climate tech. 5) Protein engineering: Antibodies, enzymes, and therapeutic proteins can now be designed at scale. What does this lead to next? The downstream effects are enormous: AI-designed drugs biologics acceleration rare disease therapeutics climate biotech enzymes protein-based materials synthetic biology platforms faster vaccine design Source: @Stanford

@snowjacuzzi Yes, it is

🤔 interesting

🏁🛰️ 🇰🇷🇺🇸🇨🇳🇸🇬 5g ai biotech alignment 🇺🇸🇰🇷 K-Moonshot : Biotech + Geoengineering 📍DeepMind launched its AI Campus in Seoul April 27th, granting SNU and KAIST direct access to AlphaGenome and AlphaFold for biotech and climate work. 📍Under K-Moonshot, the focus stays on synthetic proteins, drug pipelines, and high-resolution modeling for heatwaves and energy grids. 📍The partnership positions Seoul as an open talent hub—sometimes called Seoul-washing—drawing global researchers into a trusted, state-aligned ecosystem. 📌 KAIST is CALTECH or Standford in Korea. SNU, like the Ivy League schools that became Globalist pipeline for talent grooming, has become a cover for leftist nursery cradle of intellectual laundry and propagandistic scholarship funding. 🇸🇬🇨🇳 Singapore-washing as Exit strategy 📍Manus Parallel: The same day (April 27, 2026) China’s NDRC blocked Meta’s $2B deal for Manus. After attempting to distance itself from Chinese R&D roots, the agentic AI project faced full unwind orders, founder exit bans, and state IP claims. 📍Talent Dynamics: Seoul enables fluid exchanges and open model access. Beijing enforces tighter controls. Comparison or concern ? Is Seoul, a new talent washing hub? 🏁🛰️ Whether or not SNU and KAIST become a talent laundry pool, the correlation on protein mapping/alterning biotech with 5G aerospace race (albeit without datacenter element) on climate surveillance system agenda finds new pumping station : the corrupt government backup. 👩🏻‍💻🔎A decade after AlphaGo, South Korea’s quiet integration of frontier AI contrasts sharply with one high-profile acquisition that didn’t survive the same week. Investing motivator postings can be spotted on Pfizer, Bayer, and other top pharmaceutical investment firm gene tech but what is more interesting is the agentware.

The fourth chart explains why the market became so large in the first place. U.S. obesity rates rose from roughly 23% in the late 1980s and early 1990s to more than 42% by 2017–2018. Even though the latest reading shows a slight decline to around 40%, the long-term trend is still very clear. The obesity problem expanded steadily for decades. That means GLP-1 demand is not emerging in a small niche market. These companies are targeting one of the largest long-term public health trends in developed economies. From an economic perspective, this helps explain why pharmaceutical companies, investors, insurers, and governments are all paying enormous attention to this sector right now.

The third chart reinforces that point even more clearly. Roughly half of U.S. adults say they hear about weight-loss drugs extremely or very often. Only 19% say they rarely or never hear about them. That is remarkable for a pharmaceutical product category. Usually, drug discussions remain relatively specialized. But GLP-1 therapies have crossed into: mainstream media, social media, celebrity culture, healthcare debates, and even financial markets. This is one reason investors increasingly see obesity drugs not simply as medicines, but as broad economic platforms capable of reshaping: food consumption, healthcare spending, insurance systems, and consumer behavior

+ This is not always an easy type of data to calculate because indirect value-added trade flows are often hidden inside complex supply chains !!! Indirect value-added basically means: Chinese components, materials, or industrial inputs first move to another country, are processed or assembled there, and then finally enter the US market through that country. But one thing stands out clearly here: Mexico’s role as an intermediate hub for Chinese value-added exports to the US has grown noticeably in recent years. Mexico stands out clearly, but ASEAN economies also appear to be playing a growing intermediary role. The EU remains part of the chain as well, while the “Rest of World” category still accounts for a very large share overall. In other words, part of what appears as “Mexican exports” to the US may increasingly contain Chinese industrial value added somewhere inside the production chain. Source: @rhodium_group, Exceptional research by @CBoullenois, Malcolm Black and Alessia Caruso

. Mexico is becoming more than a “nearshoring” story The chart shows that a large share of shipments entering Mexico from China are linked either directly or indirectly to Chinese firms, especially in: smartphones, electronics components, and automotive supply chains. Some interesting patterns stand out: In smartphones and wireless equipment, Chinese-linked firms such as ZTE and Honor are already major importers inside Mexico. In electronics and data-processing equipment, Lenovo appears among the largest shipment receivers, showing how assembly and tech manufacturing networks are increasingly routed through Mexico. In automotive components, firms connected to SAIC Motor and CFMOTO stand out with very large shipment volumes. This matters because many discussions frame Mexico mainly as an alternative to China. But the data suggests something more complex is happening: In several sectors, Mexico is also becoming a manufacturing and logistics extension of Chinese industrial networks targeting the North American market. In other words, part of the supply chain may move geographically from China to Mexico, while ownership structures, intermediate inputs, engineering capability, or upstream suppliers remain heavily tied to Chinese firms ! That is why simple “China vs Mexico” narratives often miss how interconnected these production systems actually are. Source: @rhodium_group by @CBoullenois, Malcolm Black and Alessia Caruso

@UnspokenSudan Thank you very much, I hope peace spreads everywhere

Follow this account, they are producing some amazing insights!

Chinese companies’ motto this year increasingly looks like this: ""China’s carmakers chase ‘Yaris moment’ to ignite overseas growth"" Will Chinese automakers be able to achieve a global momentum similar to what Toyota once achieved with the Yaris? We will see but, it seems that this will be a difficult year for companies outside China: At first, many Chinese firms believed barriers could eventually be placed on exports directly from China. As a result, they began building overseas factories, operational hubs, and localized supply chains. The shift from “Made in China” to “Made by China” may become one of the biggest industrial stories of this decade... The next question may become: What exactly will count as “Chinese production” in the future?

Chinese companies are becoming more globalized The chart shows that overseas revenues of listed companies in China have been rising steadily over the past decade. A few important trends stand out: Overseas revenue almost tripled between 2015 and 2024. Domestic revenue also continued growing strongly, meaning international expansion happened alongside continued growth inside China rather than replacing it. The overseas share of total revenue increased from around 16% to more than 21%. This may not sound extremely high at first glance, but for an economy as large as China’s, even a few percentage points represent massive external commercial exposure. Another important point is timing. The acceleration became much stronger after 2020–2021, especially as Chinese firms expanded in: EVs, batteries, electronics, appliances, industrial machinery, and renewable energy equipment. The data also suggests that Chinese firms are increasingly relying on global demand to sustain scale economies and manufacturing utilization. In many ways, this resembles earlier export-globalization phases seen in Japan and South Korea, but the expansion is happening across far more sectors simultaneously and at a much larger industrial scale ! It also helps explain why discussions around tariffs, supply-chain diversification, and “de-risking” have become more intense globally. As Chinese firms internationalize, the distinction between domestic production and global industrial ecosystems becomes increasingly blurred.

@ManuelMCruz2 I really wish I could, but teleportation has not been invented yet :) I am in Turkiye, You would be very welcome as well But if I ever have the chance to come, I will let you know. For now, let’s continue the discussion from here

@World_Data_A We'll see. If you ever want to, I'm up to meet at a cafe and discuss about this topics.

Corea del Sur no construyó un único “Silicon Valley” biotech. Construyó una red nacional de clusters especializados. Cada ciudad cumple una función distinta dentro de la cadena de valor: investigación, manufactura, dispositivos médicos, producción industrial, comercialización.

You are absolutely right. Sometimes politicians manipulate the public very effectively. People often prefer believing comforting narratives rather than difficult realities, because accepting reality usually requires effort I think , one of the hardest things in the world is simply thinking critically. That may also be one reason why many people tend to idolize those who think on their behalf.

@World_Data_A What you point out about industrial policy is fundamental, yet 🇦🇷’s debate systematically refuses to understand it. Ironically, the most reactionary sectors constantly invoke 🇰🇷 while pretending its development came simply from “low wages,” instead of massive state coordination

@ManuelMCruz2 I am confident that Argentina will eventually manage to fix many of the things that have been going wrong

@World_Data_A Argentina is not trapped by geography or resource scarcity. I can hardly think of another case where both right and left developed such powerful reactionary forces against sustained economic development.

+ This structure in the pharmaceutical industry is truly a rare case to see at this scale.

. Among the most striking charts I have shared recently Charts by @colinterry, David Chapman, @bhangrajay and Kevin Dondarski These charts show how a surprisingly small number of drugs are gaining enormous economic power within the pharmaceutical industry. The first chart is especially revealing. A growing share of projected late-stage pipeline revenue for the world’s top 20 pharma companies is now coming from “blockbuster drugs” , medicines with extremely large sales potential. The share rises from around 51% in 2020 to nearly 70% in 2025. In other words, the industry is becoming increasingly dependent on a handful of very large drugs. This has several important implications: * successful drugs can dramatically reshape company valuations, * R&D risk becomes more concentrated, * patent battles become more important, * pricing power becomes increasingly concentrated in a small number of firms. Source: @Deloitte

@ManuelMCruz2 Thanks for the clarification

@World_Data_A Regarding LatAm integration, geography is the core problem Functionally, the region is fragmented as SEA, but islands are easier to connect because maritime transport is extremely efficient LatAm is split by mountain ranges, jungles, huge distances & sparsely populated territory

The second chart is even more interesting The green bars show how much each therapy area contributes to total industry value. The blue bars show how much of that therapy area’s value is controlled by just the top three companies. That is why the blue bars are often much longer. For example, obesity represents roughly 28% of total value creation, but almost all of that value is concentrated among a very small number of companies. Similar concentration patterns can also be seen in: endocrine & metabolism, infectious disease, genito-urinary, and musculoskeletal therapies. The broader message from these charts is very important: The pharmaceutical industry may look extremely large and diversified from the outside, but economic power is increasingly concentrating around a small number of therapies and a small number of firms.