Results 81 to 90 of about 173,284 (257)

Deciphering transcriptional plasticity in pancreatic ductal adenocarcinoma reveals alterations in sensory neuron innervation

Molecular Oncology, EarlyView.
Pancreatic sensory neurons innervating healthy and PDAC tissue were retrogradely labeled and profiled by single‐cell RNA sequencing. Tumor‐associated innervation showed a dominant neurofilament‐positive subtype, altered mitochondrial gene signatures, and reduced non‐peptidergic neurons.
Elena Genova, Michele Montrone, Uday Rangaswamy, Francesco Diversi, Irene Schiavo, Denise Ferrarini, Roberta Di Florio, Irene Longo, Michele Coscia, Nicola Zamboni, Giorgia Demontis, Lisa Veghini, Vincenzo Corbo, Remo Sanges, Paul Heppenstall   +14 more
wiley   +1 more source

Asymmetry in Low-Rank Adapters of Foundation Models

CoRR
Parameter-efficient fine-tuning optimizes large, pre-trained foundation models by updating a subset of parameters; in this class, Low-Rank Adaptation (LoRA) is particularly effective. Inspired by an effort to investigate the different roles of LoRA matrices during fine-tuning, this paper characterizes and leverages unexpected asymmetry in the ...
Jiacheng Zhu, Kristjan H. Greenewald, Kimia Nadjahi, Haitz Sáez de Ocáriz Borde, Rickard Brüel Gabrielsson, Leshem Choshen, Marzyeh Ghassemi, Mikhail Yurochkin, Justin Solomon 0001   +8 more
openaire   +3 more sources

CCDC80 suppresses high‐grade serous ovarian cancer migration via negative regulation of B7‐H3

Molecular Oncology, EarlyView.
PAX8 is a lineage‐specific master regulator of transcription in high‐grade serous ovarian cancer (HGSC) progression. We show for the first time that PAX8 facilitates proliferation and metastasis by repressing the cell autonomous tumor suppressor CCDC80 and inducing B7‐H3 expression.
Aya Saleh, Nitzan Medina‐Itzhaki, Milena Chekov, Eden Gal‐Swisa, Roba Gabesh‐Wahabi, Inbar Savyon, Inna Naroditsky, Hilary A. Kenny, Basem Fares, Lina Korsensky, Einav Girsh, Liron Berger, Ruth Perets   +12 more
wiley   +1 more source

Low-Rank Adapting Models for Sparse Autoencoders

CoRR
Sparse autoencoders (SAEs) decompose language model representations into a sparse set of linear latent vectors. Recent works have improved SAEs using language model gradients, but these techniques require many expensive backward passes during training and still cause a significant increase in cross entropy loss when SAE reconstructions are inserted ...
Matthew Chen, Joshua Engels, Max Tegmark
openaire   +3 more sources

CD47 promotes mitogen‐activated protein kinase and epithelial‐to‐mesenchymal transition molecular programs to drive prometastatic phenotypes in non‐small cell lung cancer

Molecular Oncology, EarlyView.
Beyond its role in immune evasion, this study identified that CD47 drives tumor‐intrinsic signaling in non‐small cell lung cancer (NSCLC). Transcriptomic profiling and functional studies revealed that CD47 regulates cell adhesion, migration, and metastasis through an ERK–EMT signaling axis.
Asa P.Y. Lau, Lilian G. Zhai, Ryunosuke Hoshi, Zackary Rousseau, Abraam Zakhary, Yin Fang Wu, Rola M. Saleeb, Heyu Ni, Kelsie L. Thu   +8 more
wiley   +1 more source

The Quest for Winning Tickets in Low-Rank Adapters

CoRR
The Lottery Ticket Hypothesis (LTH) suggests that over-parameterized neural networks contain sparse subnetworks ("winning tickets") capable of matching full model performance when trained from scratch. With the growing reliance on fine-tuning large pretrained models, we investigate whether LTH extends to parameter-efficient fine-tuning (PEFT ...
Hamed Damirchi, Cristian Rodriguez Opazo, Ehsan Abbasnejad, Zhen Zhang 0008, Javen Shi   +4 more
openaire   +2 more sources

KDM7A and KDM1A inhibition suppresses tumour promoting pathways in prostate cancer

Molecular Oncology, EarlyView.
Treatment resistance is a major challenge for patients with advanced prostate cancer. This study examined an alternative approach to target the major prostate cancer‐promoting pathway by targeting epigenetic factors, whose levels are higher in tumours.
Jennie N Jeyapalan, Veronika M Metzler, Simone de Brot, Corinne L Woodcock, Anna E Harris, Jennifer Lothion‐Roy, Emeli M Nilsson, Atara Ntekim, Michael S Toss, Jenny L Persson, Francesca Khani, Brian D Robinson, Lorraine J Gudas, Emad Rakha, David M Heery, Catrin S Rutland, Nigel P Mongan   +16 more
wiley   +1 more source

Provable Meta-Learning with Low-Rank Adaptations

CoRR
The power of foundation models (FMs) lies in their capacity to learn highly expressive representations that can be adapted to a broad spectrum of tasks. However, these pretrained models require additional training stages to become effective for downstream applications.
Jacob L. Block, Sundararajan Srinivasan, Liam Collins, Aryan Mokhtari, Sanjay Shakkottai   +4 more
openaire   +2 more sources

Heterozygous loss‐of‐function alleles associate the conserved 3′‐5′ exoribonuclease EXOSC10 with hypersensitivity to the anticancer drug 5‐fluorouracil

Molecular Oncology, EarlyView.
EXOSC10, an essential nuclear RNA exosome‐associated 3′‐5′ exoribonuclease, is inhibited by the anticancer drug 5‐fluorouracil (5‐FU), and EXOSC10 depletion increases 5‐FU sensitivity. The colon‐cancer variant EXOSC10S402T, located in a proteolysis motif, is stable and nuclear but nonfunctional in vivo.
Radhika Sain, Yann Le Page, Frédéric Percevault, Emmanuelle Becker, Luc Negroni, Almudena Fernandez, Marta Cantero, Diego Muñoz‐Santos, Lluís Montoliu, Cyrille Garnier, Michael Primig   +10 more
wiley   +1 more source

dEBORA: Efficient Bilevel Optimization-based low-Rank Adaptation [PDF]

Low-rank adaptation methods are a popular approach for parameter-efficient fine-tuning of large-scale neural networks. However, selecting the optimal rank for each layer remains a challenging problem that significantly affects both performance and ...
Zangrando, Emanuele, Rinaldi, Francesco, Tudisco, Francesco, Venturini, Sara   +3 more
core