Honoring the legacy of Nobel Laureate #RobertWilliamHolley on his #BirthAnniversary at the #NobelPrizeGallery! From sequencing #tRNA to inspiring young minds—students at @RSCBhavnagar explored genetics hands-on by building #DNAmodels and decoding life’s blueprint 🧬

Celebrated the #birthanniversary of Nobel Laureate #RobertWilliamHolley at #NobelPrizeGallery He was the first to sequence #tRNA, bridging the gap between DNA and life itself. students are honoring his legacy @RSCBhavnagar by building their own #DNAmodels and cracking the code!

"What Tyshawn does . . ." For @GQMagazine Edited by @PAPPADEMAS Thank you TJ, @dnamodels @MTA gq.com/story/tyshawn-jones-g…

Nucleotide Dependency Analysis of DNA Language Models Reveals Genomic Functional Elements This new study introduces the concept of nucleotide dependencies, a novel and interpretable metric derived from DNA language models (DNA LMs). This method quantifies how changes at one genomic position influence the probabilities of nucleotides at other positions, providing a powerful tool for deciphering functional elements. The researchers generated genome-wide maps of these pairwise nucleotide dependencies across animal, fungal, and bacterial species. These maps were shown to be highly effective in indicating the deleteriousness of human genetic variants, outperforming both traditional sequence alignment and direct DNA LM reconstruction methods. A significant innovation is the identification of regulatory elements, such as transcription factor binding sites, which appear as dense blocks within these dependency maps. This approach enables their systematic and accurate identification, performing comparably to models trained on experimental binding data, despite the dependency analysis being entirely unsupervised. Beyond DNA, nucleotide dependencies also precisely highlight bases that interact within RNA structures, including complex pseudoknots and tertiary contacts. This capability led to the exciting discovery and experimental validation of four previously unknown RNA structures in Escherichia coli. Furthermore, the dependency maps serve as a diagnostic tool for evaluating DNA LM architectures and training data selection strategies. The analysis revealed critical limitations in certain models and underscored the importance of training on multi-species genomic data for robust learning of infrequent but highly conserved functional elements. Overall, this work establishes nucleotide dependency analysis as a promising new avenue for discovering, characterizing, and understanding functional elements and their intricate interactions throughout genomes. 💻Code: github.com/gagneurlab/depend… 📜Paper: biorxiv.org/content/10.1101/… #ComputationalBiology #Genomics #DNAModels #RNAStructure #Bioinformatics #GenomicFunctionalElements #MachineLearning

Nucleotide Dependency Analysis of DNA Language Models Reveals Genomic Functional Elements This new study introduces the concept of nucleotide dependencies, a novel and interpretable metric derived from DNA language models (DNA LMs). This method quantifies how changes at one genomic position influence the probabilities of nucleotides at other positions, providing a powerful tool for deciphering functional elements. The researchers generated genome-wide maps of these pairwise nucleotide dependencies across animal, fungal, and bacterial species. These maps were shown to be highly effective in indicating the deleteriousness of human genetic variants, outperforming both traditional sequence alignment and direct DNA LM reconstruction methods. A significant innovation is the identification of regulatory elements, such as transcription factor binding sites, which appear as dense blocks within these dependency maps. This approach enables their systematic and accurate identification, performing comparably to models trained on experimental binding data, despite the dependency analysis being entirely unsupervised. Beyond DNA, nucleotide dependencies also precisely highlight bases that interact within RNA structures, including complex pseudoknots and tertiary contacts. This capability led to the exciting discovery and experimental validation of four previously unknown RNA structures in Escherichia coli. Furthermore, the dependency maps serve as a diagnostic tool for evaluating DNA LM architectures and training data selection strategies. The analysis revealed critical limitations in certain models and underscored the importance of training on multi-species genomic data for robust learning of infrequent but highly conserved functional elements. Overall, this work establishes nucleotide dependency analysis as a promising new avenue for discovering, characterizing, and understanding functional elements and their intricate interactions throughout genomes. 💻Code: github.com/gagneurlab/depend… 📜Paper: biorxiv.org/content/10.1101/… #ComputationalBiology #Genomics #DNAModels #RNAStructure #Bioinformatics #GenomicFunctionalElements #MachineLearning

BMFM-DNA: A SNP-aware DNA foundation model to capture variant effects 1. Researchers have developed BMFM-DNA, a groundbreaking approach to DNA language models that directly integrates Single Nucleotide Polymorphisms (SNPs) and other sequence variations during pre-training. This innovation addresses a key limitation of previous models that often overlooked the crucial biological impact of genomic variations. 2. The team pre-trained two Biomedical Foundation Models (BMFM) using ModernBERT: BMFM-DNA-REF, trained on reference genome sequences, and BMFM-DNA-SNP, which incorporates a novel representation scheme to encode sequence variations. This dual approach allowed for comprehensive evaluation of variation integration. 3. A significant innovation for BMFM-DNA-SNP involves mapping genetic variants from the dbSNP database, including SNPs, insertions, and deletions, to unique Chinese characters. This unique encoding implicitly introduces multiple nucleotide possibilities at a single genomic position, thereby expanding the effective pre-training DNA sample space and revealing hidden patterns of variation distribution. 4. Experiments showed that integrating sequence variations into these DNA language models leads to notable improvements across various fine-tuning tasks, demonstrating their enhanced ability to capture complex biological functions that are influenced by genomic differences. 5. The underlying architecture, ModernBERT, is a modernized encoder-only transformer model designed for improved efficiency and performance, particularly with longer sequence lengths. It incorporates features like Rotary Positional Embeddings (RoPE) and FlashAttention. 6. To support further research and community contributions, the models and the code for reproducing the results have been publicly released through HuggingFace and GitHub. A comprehensive software package, bmfm-multi-omic, is also available for pre-training, finetuning, and benchmarking genomic foundation models. 💻Code: github.com/BiomedSciAI/biome… 📜Paper: arxiv.org/pdf/2507.05265v1 #ComputationalBiology #Genomics #FoundationModels #SNPs #DNAModels #Bioinformatics #AI #MachineLearning

BMFM-DNA: A SNP-aware DNA foundation model to capture variant effects 1. Researchers have developed BMFM-DNA, a groundbreaking approach to DNA language models that directly integrates Single Nucleotide Polymorphisms (SNPs) and other sequence variations during pre-training. This innovation addresses a key limitation of previous models that often overlooked the crucial biological impact of genomic variations. 2. The team pre-trained two Biomedical Foundation Models (BMFM) using ModernBERT: BMFM-DNA-REF, trained on reference genome sequences, and BMFM-DNA-SNP, which incorporates a novel representation scheme to encode sequence variations. This dual approach allowed for comprehensive evaluation of variation integration. 3. A significant innovation for BMFM-DNA-SNP involves mapping genetic variants from the dbSNP database, including SNPs, insertions, and deletions, to unique Chinese characters. This unique encoding implicitly introduces multiple nucleotide possibilities at a single genomic position, thereby expanding the effective pre-training DNA sample space and revealing hidden patterns of variation distribution. 4. Experiments showed that integrating sequence variations into these DNA language models leads to notable improvements across various fine-tuning tasks, demonstrating their enhanced ability to capture complex biological functions that are influenced by genomic differences. 5. The underlying architecture, ModernBERT, is a modernized encoder-only transformer model designed for improved efficiency and performance, particularly with longer sequence lengths. It incorporates features like Rotary Positional Embeddings (RoPE) and FlashAttention. 6. To support further research and community contributions, the models and the code for reproducing the results have been publicly released through HuggingFace and GitHub. A comprehensive software package, bmfm-multi-omic, is also available for pre-training, finetuning, and benchmarking genomic foundation models. 💻Code: github.com/BiomedSciAI/biome… 📜Paper: arxiv.org/pdf/2507.05265v1 #ComputationalBiology #Genomics #FoundationModels #SNPs #DNAModels #Bioinformatics #AI #MachineLearning

Name: Tate Short Role: Model Born: 1998 From: Victoria, Australia Height: 6ft 2¾in [190cm] Hair: Dark Blonde Eyes: Green-Hazel IG: @tatelshort #tateshort Agencies: @precisionmgmt @viviensmodelmgmt @dnamodels

Name: Jack Kalani Pilila’au Role: Model Born: 2000 From: Arizona, USA Height: 6ft 4in [193cm] Hair: Brown Eyes: Hazel IG: @jack.kalani #jackkalani #jackkalanipililaau Agencies: @wilhelminamodels @nextmodels @dnamodels @whynotmodels

REPOST @dnamodels I’m so proud of my special daughter Kirsty Hume! For always being true to her extraordinary self, here at @CHANEL and everywhere – I love her so much! 👏💜🧡💋

Summer lovin’ #GIANTmicrobes style 😂🧬 Perfect for #backtoschool, college gifts and just because! Shop our #DNA, #RNA and more #Genetics here: giantmicrobes.com/us/main/ge… #microbes #collegestudents #genetics #epigenetics #dnamodels #biologymemes #biologyjokes #science #biology

@dnamodels @elitemodels @vivamodelmgmt @FordModels come and have a look at a beautiful face card

@Wilhelmina @LAModelsAgency @dnamodels get him

I need one of you to sign @ShileseJ ASAP! The world needs to see that face card in beauty campaigns! @FordModels @IMGmodels @dnamodels @VNYmodels @Wilhelmina @MarilynAgencyNY

@StormModels @JagModels @nextmanagementx @dnamodels @LAModelsAgency one shoot is all that I ask for you to collaborate with @yolandamukondi & the rest you will take it from there YOLANDA MONYAI ONCE SAID #YolandaMonyai #YolandaMukondeleli #YolandaMokondeleli

#Xcoo regularly organize #XcooHackDay, and have some drinks and snacks while getting hands-on experience in our office. Our third event was titled "Let's Assemble a DNA Molecule Model". We made a hanging DNA molecule model assembly kit to create #DNAmodels .

@elitemodels @Wilhelmina @dnamodels @FordModels @NextModels 📸

"Where are you in life? It's not really vertically, it's more horizontally. It's not like this is higher and this is lower. It's just... what spectrum are you on?" @JuliaBergshoeff Julia Bergshoeff filmed by Andy Swartz for What’s in your dna @dnamodels > instagram.com/reel/Cri2wiNgb…

Not feeling the gearhead Prosi look at all on her.