Sri Nandan Gondi

0 followers · 251 views

on the atlas · 7

highlights · 23

more comprehensively and accurately predicts how single variants or mutations in human DNA sequences impact a wide range of biological processes regulating genes.
AlphaGenome: AI for better understanding the genome — Google DeepMind
Model overview. AlphaGenome processes 1 Mb of DNA sequences and species identity (human/mouse) to predict 5,930 human or 1,128 mouse genome tracks across diverse cell types and 11 output types at specific resolutions (far right). Computation leverages sequence parallelism, breaking the 1 Mb of DNA sequence into 131-kb chunks processed across devices. The core architecture features a U-Net-style design comprising an encoder (downsampling the sequence), transformers with inter-device communication and a decoder (upsampling), which feed into task-specific output heads at their respective resoluti…
Advancing regulatory variant effect prediction with AlphaGenome | Nature
For example, we used AlphaGenome to investigate the potential mechanism of a cancer-associated mutation. In an existing study of patients with T-cell acute lymphoblastic leukemia (T-ALL), researchers observed mutations at particular locations in the genome. Using AlphaGenome, we predicted that the mutations would activate a nearby gene called TAL1 by introducing a MYB DNA binding motif, which replicated the known disease mechanism and highlighted AlphaGenome’s ability to link specific non-coding variants to disease genes.
AlphaGenome: AI for better understanding the genome — Google DeepMind
To facilitate broader use, we provide tools for making genome track and variant effect predictions from sequence.
Advancing regulatory variant effect prediction with AlphaGenome | Nature
The AlphaGenome architecture uses convolutional layers to initially detect short patterns in the genome sequence, transformers to communicate information across all positions in the sequence, and a final series of layers to turn the detected patterns into predictions for different modalities. During training, this computation is distributed across multiple interconnected Tensor Processing Units (TPUs) for a single sequence.
AlphaGenome: AI for better understanding the genome — Google DeepMind
Our AlphaGenome model takes a long DNA sequence as input — up to 1 million letters, also known as base-pairs — and predicts thousands of molecular properties characterising its regulatory activity. It can also score the effects of genetic variants or mutations by comparing predictions of mutated sequences with unmutated ones.
AlphaGenome: AI for better understanding the genome — Google DeepMind
But deciphering how the genome’s instructions are read at the molecular level — and what happens when a small DNA variation occurs — is still one of biology’s greatest mysteries.
AlphaGenome: AI for better understanding the genome — Google DeepMind
This architectural choice of diffusion means that Atomwell is not only very good at predicting properties of small molecules and proteins for downstream tasks (via embeddings), but also excellent at conditional generation of proteins/small molecules given a motif/part of a sequence.
Defeating pests with AI models: Our first-principles approach
Instead of using human knowledge to generate proteins and small molecules on the basis of known or hypothesized properties, we leverage foundational models trained on massive amounts of data.
Defeating pests with AI models: Our first-principles approach
‘general methods that leverage computation are ultimately the most effective, and by a large margin’
Defeating pests with AI models: Our first-principles approach
At the same time, agricultural innovation has stagnated, with fewer than 20 new active ingredients approved in the past decade — down from 40 in the decade before that.
Defeating pests with AI models: Our first-principles approach
As pests evolve and develop resistance, farmers are forced to use increasing amounts of pesticides to get the same results, damaging the environment and leading to even more pest resistance
Defeating pests with AI models: Our first-principles approach
Pesticide usage has doubled per acre over the last three decades
Defeating pests with AI models: Our first-principles approach
We support this hypothesis by showing in-vitro data for pesticides discovered by our model
Defeating pests with AI models: Our first-principles approach
We believe that foundational models have become powerful tools for predicting these different properties, and hence can be leveraged with generative pipelines to effectively find pesticides.
Defeating pests with AI models: Our first-principles approach
In general, we view pesticide discovery as a search space problem, where we’re trying to optimize for multiple different (sometimes conflicting) properties of small molecules and peptides.
Defeating pests with AI models: Our first-principles approach
In 1887, for example, 10,000 sugar plantation workers in Louisiana, almost all of them Black and members of the Knights, struck for higher wages and an end to payment in scrip.
Community and Conflict: Working People Respond to Industrial Capitalism, 1877-1893 | Who Built America? Working People and the Nation’s History | Who Built America?
he Founding, the Civil War, the New Deal. The holy trinity of the American political tradition.
Our Forgotten Labor Revolution
Working people, in response, shook their collective fist at the growing visibility of unbridled privilege, especially in the sixteen-year period framed by the railroad strikes of 1877 and the depression of 1893. Workers joined together in the Knights of Labor, the eight-hour movement, and the craft unions. They struck not only for higher wages, but also to express solidarity with their fellow workers.
Community and Conflict: Working People Respond to Industrial Capitalism, 1877-1893 | Who Built America? Working People and the Nation’s History | Who Built America?
Humorists Mark Twain and Charles Dudley Warner’s novel The Gilded Age (1874) satirized the politics and values of the post–Civil War boom years: “Get rich . . . dishonestly if we can, honestly if we must.” Historians later adopted their title to describe the materialism and superficiality of the late nineteenth century.
Community and Conflict: Working People Respond to Industrial Capitalism, 1877-1893 | Who Built America? Working People and the Nation’s History | Who Built America?
Bahdanau et al proposed an attention mechanism in their 2014 paper, “Neural Machine Translation by Jointly Learning to Align and Translate,” to improve communication between the encoder and decoder and remove that information bottleneck. Instead of passing along only the final hidden state of the encoder—the context vector—to the decoder, their model passed every encoder hidden state to the decoder. The attention mechanism itself was used to determine which hidden state—that is, which word in the original sentence—was most relevant at each translation step performed by the decoder. “This frees…
What is an attention mechanism? | IBM
The 2017 invention of transformer models, powered solely by attention, eventually made RNNs all but obsolete for NLP.
What is an attention mechanism? | IBM
Mathematically speaking, an attention mechanism computes attention weights that reflect the relative importance of each part of an input sequence to the task at hand.
What is an attention mechanism? | IBM