Editorial 1 : On the Hindu Kush Himalayas snow update
Context
The Ganga river basin — India’s largest — reached a record low snow persistence in 2024, the Hindu Kush Himalaya snow update of the International Centre for Integrated Mountain Development (ICIMOD) has reported.
What is snow persistence?
- Snow persistence is the fraction of time snow is on the ground. When this snow melts, it provides water to people and ecosystems.
- In the river basins of the Hindu Kush Himalaya (HKH), snowmelt is the biggest source of water in the streams.
- Overall, it contributes 23% of the runoff to the region’s 12 major river basins every year.
- The HKH mountains extend around 3,500 km over eight countries — Afghanistan, Bangladesh, Bhutan, China, India, Nepal, Myanmar, and Pakistan.
- These mountains are also called the “water towers of Asia” because they are the origins of 10 crucial river systems on the continent — Amu Darya, Indus, Ganga, Brahmaputra, Irrawaddy, Salween, Mekong, Yangtse, Yellow river, and Tarim.
- These river basins provide water to almost one-fourth of the world’s population and are a significant freshwater source for 240 million people in the HKH region.
The Report
- The authors of the 2024 HKH snow update analysed data from 2003 to 2024 and found significant fluctuations in snow persistence between November and April every year, when snow accumulates above ground.
- In India, snow persistence in the Ganga, the Brahmaputra, and the Indus river basins dropped significantly in 2024.
- Similarly, snow persistence in the Brahmaputra basin was 14.6% below normal in 2024. It was worse in 2021, when the average persistence was 15.5% below normal.
- In the Indus river basin, snow persistence fell 23.3% below normal this year although this was offset by excesses in parts of the lower altitudes.
- Outside India, the basin of the Amu Darya river — which flows through Central Asia — recorded its lowest snow persistence in 2024: 28.2% below normal.
- The figure for the Helmand river, an important source of drinking water for Iran and Afghanistan, was almost 32% below normal in 2024, beating a record set in 2018.
- Persistence in the part where the Mekong river originates in the Himalaya was only slightly below normal this time. (This river’s delta is Vietnam’s “rice bowl”.)
The Reasons
- The primary reason for the lower persistence in 2024 was weak western disturbances.
- Due to changing climate and global warming, this pattern is becoming increasingly unstable.
- Although the exact mechanisms are not fully understood, global warming is thought to exacerbate prolonged and intense La Niña–El Niño conditions.
- These phases of a recurring climate pattern across the tropical Pacific Ocean significantly influence global weather patterns, including western disturbances.
- Western disturbances are low-pressure systems that originate over the Mediterranean Sea, the Caspian, and the Black Seas and bring rain and snow to the HKH region in winter.
- The region where these storms originate experienced persistently high sea-surface temperatures.
- This disruption weakened and delayed the arrival of the western disturbance, resulting in reduced winter precipitation and snowfall in the HKH region.
- The pattern of high temperatures and altered weather systems explains both the record low snow persistence in 2024 and similar historical records.
What explains higher snow persistence?
- The persistence of snow in China’s Yellow River basin exceeded the normal value by 20.2% in 2024.
- The Yellow river basin is an area where the East Asian winter monsoon brings cold, dry air from Siberia and Mongolia.
- When this cold air mass interacts with moist air from other regions, particularly the Pacific Ocean, it can result in snowfall over the higher altitudes of the upper Yellow River basin.
- When the cold air from the east Asian winter monsoon systems interacts with moist air masses from the Pacific Ocean, it can result in snowfall at higher elevations in the eastern Himalaya.
What about India?
- Snow persisting on the ground is important for the Ganga river basin because its melt contributes to 10.3% of the latter’s water, versus 3.1% from glacier melts.
- In the Brahmaputra and the Indus basins as well, snowmelt brings 13.2% and around 40% of the water, respectively, versus 1.8% and 5% from glaciers.
- Lower snow in 2024 may affect water availability, particularly and most importantly in the Indus basin, if there is less rainfall in the early season.
- In the long term, experts say, reforestation with native tree species can help the ground retain more snow.
- Better weather forecasting and early warning systems can also help local communities prepare for impending water stress. “
- Improving water infrastructure and developing policies for protecting areas receiving snowfall are important for long-term change.
- Communities involvement in local, national level decision-making and promoting regional cooperation are vital for comprehensive solutions for the sustainability of snow.
Conclusion
- There is a need to reduce emissions, which would mitigate increasing sea-surface and ground temperatures, both of which lower the persistence of snow. The key work for all of us concerned about a liveable future on the earth is to build the political will for our government representatives and business leaders to cut the cord on dirty fossil energy consumption and production, especially among G-20 countries, which account for 81% of all emissions.
Editorial 2 : How will AI that predicts protein structures change the life sciences?
Introduction
Proteins are one of the most important molecules of life, with almost every biological function from birth to death being regulated by them in some way. Each protein is made up of a string of smaller building blocks called amino acids, which contain all the information to transform proteins — from a single sequence to a folded, functional 3D structure.
Answers ex machina
- The steps a protein takes to go from its straight form to its final form are too many to count and too hard to follow, leaving the question of how every protein folds — the famous protein-folding problem — unanswered.
- If we want to understand the molecular basis of how cells work, how organisms work, how life works, we need to understand how proteins get their shape.
- Things changed when Google DeepMind’s protein-structure prediction software AlphaFold burst onto the scene in 2020.
- The highly improved AlphaFold 2 was introduced in 2021.
- AlphaFold uses machine learning and artificial intelligence (AI) to accurately predict protein structures from an amino acid sequence, seemingly solving the protein-folding problem without learning any of the deeper physical principles that drive this biological process.
- Now, in a Nature paper published in May 2024, scientists at DeepMind led by John Jumper introduced AlphaFold 3, building on its predecessors with even more transformative capabilities.
- AlphaFold 3 can predict protein-protein interactions as well as the structures of other molecules like DNA and RNA, along with the interactions of proteins with all these other compounds.
Democratising research
- AlphaFold 2 predicted the structure of proteins with revolutionary levels of accuracy.
- AlphaFold 3 is even more accurate for proteins, but can also predict the structure of DNA, RNA, and all the other molecular components that make up biology.
- The interaction of all these biomolecules is what makes up the processes of life, so it is important to be able to predict the structure of these interactions.
- Apart from being able to give us a lot more insight into biological processes, the new AlphaFold is also more usable by scientists who aren’t experts in machine learning.
From noise to signal
- The original AlphaFold was trained on the thousands of sequences and protein structures present in the protein data bank, a giant protein repository where scientists submit experimentally determined protein structures.
- It completely ignores all the fundamental physics and thermodynamics, it’s modelling based on learning what real structures tend to look like, taking advantage of tendencies of protein structures that are too subtle for humans to realise.
- Unlike its predecessors, AlphaFold 3 uses a diffusion model, which is what image-generating software also uses.
- The model works by first training on protein structures, adding noise to the data, and then trying to de-noise it.
- This way, the model becomes able to work its way back from a noisy structure to a real protein structure. This architecture also helps AlphaFold 3 handle a much larger input dataset.
A reliability problem
- Its accuracy at predicting protein-protein interactions is also incredibly high — but not its reliability when it comes to interactions between small molecules and proteins.
- Proteins use a language of 20 amino acids whereas small molecule ligands have a much larger vocabulary.
- Greater variations in the dataset and the use of diffusion techniques can lead to the model coming up with answers that look plausible but aren’t real.
- Adding more training data can help circumvent this problem, but not entirely get rid of it.
- Nevertheless, AlphaFold 3 predicts protein structures and interactions better than other models right now.
- Academics and companies can potentially use it to find drug candidates that can bind to proteins and help cure diseases.
Conclusion
- Additionally, even though scientists are free to use the AlphaFold server to upload their protein sequences, many researchers are irked at not being able to access the model’s full code. This means they can’t play around with its nuts and bolts and modify it for specific use-cases.
- Different groups have also begun a race to crack the model’s code and make open-source versions.
