If you have been reading about GRAS and want a single page that covers the useful parts, this is it: definitions, context, how it is studied, and the questions that come up repeatedly.
Updated 2025-11-23. Numbers and descriptions here follow the published literature rather than marketing material.
Raw collagen for peptide production comes from bovine hide, porcine skin, fish skin and scales, and sometimes poultry cartilage. The material is cleaned, extracted, and treated with acid, alkali, or enzymes to break peptide bonds. Enzymatic hydrolysis using proteases allows better control of fragment size than purely chemical methods. After hydrolysis, the liquid is filtered, concentrated, and dried into a powder. Source and processing conditions influence color, odor, molecular weight distribution, and amino acid profile.
The distinction between native collagen and collagen peptides matters for behavior in water and in analytical tests. Native collagen is a rigid, triple-helical protein that is largely insoluble in cold water. Peptides lack that organized helix and dissolve readily, forming clear or slightly hazy solutions. Because hydrolysis shortens chains, viscosity falls and gelation behavior changes. The term collagen peptide does not specify a single molecular species; it describes a family of hydrolysates with variable chain lengths and properties.
Collagen peptides are short-chain proteins produced by hydrolyzing native collagen, the main structural protein in skin, bone, tendon, and cartilage. The hydrolysis step breaks the triple-helical structure and cleaves longer chains into smaller fragments. The resulting material is water-soluble and typically has an average molecular weight in the low kilodalton range. Commercial ingredients are often described as hydrolyzed collagen or collagen hydrolysate. Amino acid composition remains rich in glycine, proline, and hydroxyproline, though the ordered helical arrangement is largely lost.
Analytical testing of collagen peptides focuses on identity, purity, and molecular weight profile. Size-exclusion chromatography separates peptides by hydrodynamic volume and is often calibrated with known protein standards. Amino acid analysis after acid hydrolysis provides the compositional profile, which can confirm the collagen origin. Mass spectrometry offers detailed sequence information for individual peptides. These methods together help ensure that a product matches its specification and that batch-to-batch variability is controlled.
Dry collagen peptide powder is generally stable when kept in a sealed container away from moisture, heat, and direct sunlight. The powder is hygroscopic and can clump if exposed to humid air, so desiccant packets are sometimes included. In solution, collagen peptides are susceptible to microbial growth unless preserved or refrigerated. Prolonged exposure to high temperatures may cause aggregation or color changes. Typical storage recommendations are cool and dry conditions at ambient temperature.
Quality control for collagen peptides includes measurements of moisture content, ash, protein content, and heavy metals. Microbial limits are set to ensure food or cosmetic grade safety, and the degree of hydrolysis serves as a key process indicator. That indicator correlates with molecular weight distribution and solubility characteristics. Regulatory requirements vary by country, and some jurisdictions restrict label claims about health effects. Documentation such as certificates of analysis and safety data sheets typically accompanies commercial shipments of the material.
| Property | Value | Notes |
|---|---|---|
| Common synonyms | Hydrolyzed collagen; collagen hydrolysate | Terms used interchangeably in ingredient lists |
| Appearance | White to off-white powder | Color can vary with source and processing |
| Solubility | Freely soluble in water | Insoluble in ethanol and many organic solvents |
| Typical molecular weight | 1-10 kDa | Average often around 2-6 kDa depending on process |
| Typical storage | Dry, 15-25 °C | Protect from moisture and strong odors |
Collagen is a structural protein found in connective tissues of animals, and collagen peptides are short amino acid chains produced by hydrolyzing native collagen into smaller fragments. The hydrolysis process typically uses enzymes or acids under controlled conditions. Commercial collagen peptides often come from bovine hide, porcine skin, or fish scales. The resulting material is water-soluble and differs from intact collagen in molecular size and behavior. The term 'collagen peptide' generally refers to a mixture of peptide chains rather than a single defined molecule.
Production begins with cleaning and mincing raw collagen-rich tissues. The material undergoes pretreatment to remove non-collagenous components, followed by hydrolysis using enzymes such as pepsin or alcalase, or by acid or alkaline treatment. Reaction time, temperature, and pH influence the average molecular weight of the resulting peptides. After hydrolysis, the mixture is filtered, concentrated, and dried, often by spray drying. The final product is a powder with a characteristic amino acid profile rich in glycine, proline, and hydroxyproline.
Collagen peptides are distinguished from gelatin by their lower average molecular weight and better solubility in cold water. Gelatin forms gels upon cooling, while collagen peptides typically do not. Molecular weight distributions for commercial collagen peptides often range from about 2 to 20 kilodaltons, though exact profiles vary by manufacturer and process. Products may be sold as powders, capsules, or liquids. The term "collagen hydrolysate" is frequently used as a synonym, although labeling conventions differ across regions.
Manufacturing collagen peptides begins with collagen-rich raw materials such as bovine hide, porcine skin, fish scales, or poultry cartilage, which undergo washing, size reduction, and pretreatment to remove non-collagen proteins and fats. Extraction may use acid, alkali, or heat. Hydrolysis then breaks the collagen into smaller peptides, often with enzymes such as pepsin, papain, or alcalase. Process conditions of time, temperature, pH, and enzyme dose determine the final molecular weight distribution. After hydrolysis, the solution is filtered, concentrated, and dried into powder.
Quality testing of collagen peptides relies on several analytical methods. Molecular weight distribution is commonly measured by size-exclusion chromatography, sometimes paired with multi-angle light scattering. Amino acid composition is determined by ion-exchange chromatography or reversed-phase high-performance liquid chromatography after acid hydrolysis, while protein content is estimated by Kjeldahl or Dumas nitrogen analysis. Moisture, ash, and heavy metals are checked against specification limits. These tests help ensure consistency and detect adulteration with other proteins.
Most commercial collagen peptides derive from bovine hide, porcine skin, fish skin, or poultry cartilage, with fish sources often having lower thermal stability. Their amino acid profile is distinctive: glycine appears at roughly every third residue in the parent collagen triple helix, and proline and hydroxyproline are abundant. Collagen itself lacks tryptophan and is low in several essential amino acids, so collagen peptides are not a complete protein source. Source tissue and processing can influence peptide length, amino acid composition, color, odor, and mineral content.
Hydrolysis conditions determine the peptide size profile, which in turn affects solubility, viscosity, taste, and behavior in formulations. Products may contain free amino acids, di- and tripeptides, and larger fragments up to tens of kilodaltons. Average molecular weight is often reported, but the distribution is more informative because two materials with the same average can differ in peptide profile. Ultrafiltration, spray drying, and ion exchange may be used to standardize the final powder. The relationship between specific peptide sequences and measured effects remains an active area of study.
Plutonium-238 (238Pu or Pu-238) is a radioactive isotope of plutonium that has a half-life of 87.7 years. Plutonium-238 is a very powerful alpha emitter; as alpha particles are easily blocked, this makes the plutonium-238 isotope suitable for usage in radioisotope thermoelectric generators (RTGs) and radioisotope heater units. The density of plutonium-238 at room temperature is about 19.8 g/cc. The material will generate about 0.57 watts per gram of 238Pu. The bare sphere critical mass of metallic plutonium-238 is not precisely known, but its calculated range is between 9.04 and 10.07 kg (19.9 and 22.2 lb).
However, several academic researchers became concerned that AI was no longer pursuing its original goal of creating versatile, fully intelligent machines. Beginning around 2002, they founded the subfield of artificial general intelligence (or "AGI"), which had several well-funded institutions by the 2010s. Deep learning began to dominate industry benchmarks in 2012 and was adopted throughout the field. For many specific tasks, other methods were abandoned. Deep learning's success was based on both hardware improvements (faster computers, graphics processing units, cloud computing) and access to large amounts of data (including curated datasets, such as ImageNet). Neural-network approaches also advanced natural language processing. In 2013, word2vec introduced efficient methods for learning distributed word representations. In 2014–2015, sequence-to-sequence and attention mechanisms advanced neural machine translation. Deep learning's success led to an enormous increase in interest and funding in AI. The amount of machine learning research (measured by total publications) increased by 50% in the years 2015–2019.
His observations culminated in the Political Essay on the Island of Cuba, a comprehensive geographic study examining the island’s physical and economic conditions, as well as the social realities of slavery. Published in 1828, the work was notable for its data and its condemnation of slavery, resonating during the Latin American independence movements. If the legislation of the Antilles and the condition of the colored population does not experience some salutary change, and if discussion without action is continued, the political power may well pass into the hands of that class which holds the might of labor, the will to throw off the yoke Humboldt also analyzed demographic data. In Cumana, he had recorded 6,000 people of color among 110,000 white and Creole residents. In Havana, government archives revealed that more than 2,130,000 Africans had been forcibly transported to British Caribbean territories over the previous century. In 1806, the slave trade involved 53,000 sales in British dominions and 15,000 in the United States. Humboldt estimated that from 1670 to 1825, nearly five million Africans were brought to the West Indies, not counting deaths during the Middle Passage. He was deeply affected by the realities of slavery, expressing indignation and advocating for strict enforcement of anti-slavery laws. He hoped that anti-slavery principles would spread southward in the Americas. The journey through Cuba was shortened when Humboldt received news of Captain Baudin’s French scientific expedition, expected on the Peruvian coast within a year.
Sources: en.wikipedia.org
== Capture of Constantinople == When the Nicaean force reached the village of Selymbria, some 30 miles (48 km) west of Constantinople, they learned from some independent local farmers (thelematarioi) that the entire Latin garrison, as well as the Latin-allied Venetian fleet, was absent conducting a raid against the Nicaean island of Daphnousia in the Black Sea. Strategopoulos initially hesitated to take advantage of the situation, fearing that his small force might be destroyed if the Latin army returned too soon, and because he would exceed the provisions of the emperor's orders, but eventually decided he could not squander such a remarkable opportunity to seize the city without a prolonged siege.
==== 1.A α-type channels ==== 1.A.1 Voltage-gated ion channel superfamily 1.A.2 Inward-rectifier K+ channel family 1.A.3 Ryanodine-inositol-1,4,5-trisphosphate receptor Ca2+ channel family 1.A.4 Transient receptor potential Ca2+ channel family 1.A.5 Polycystin cation channel family 1.A.6 Epithelial Na+ channel family 1.A.7 ATP-gated P2X receptor cation channel family 1.A.8 Major intrinsic protein superfamily 1.A.9 Neurotransmitter receptor, Cys loop, ligand-gated ion channel family 1.A.10 Glutamate-gated ion channel family of neurotransmitter receptors 1.A.11 Ammonium channel transporter family 1.A.12 Intracellular chloride channel family 1.A.13 Epithelial chloride channel family 1.A.14 Testis-enhanced gene transfer family 1.A.15 Nonselective cation channel-2 family 1.A.16 Formate-nitrite transporter family 1.A.17 Calcium-dependent chloride channel family 1.A.18 Chloroplast envelope anion-channel-forming Tic110 family 1.A.19 Type A influenza virus matrix-2 channel family 1.A.20 BCL2/Adenovirus E1B-interacting protein 3 family 1.A.21 Bcl-2 family 1.A.22 Large-conductance mechanosensitive ion channel 1.A.23 Small-conductance mechanosensitive ion channel 1.A.24 Gap-junction-forming connexin family 1.A.25 Gap-junction-forming innexin family 1.A.26 Mg2+ transporter-E family 1.A.27 Phospholemman family 1.A.28 Urea transporter family 1.A.29 Urea/amide channel family 1.A.30 H+- or Na+-translocating bacterial MotAB flagellar motor/ExbBD outer-membrane transport energizer superfamily 1.A.31 Annexin family 1.A.32 Type B influenza virus NB channel family 1.A.33 Cation-channel-forming heat shock protein 70 family 1.A.34 Bacillus gap junction-like channel-forming complex family 1.A.35 CorA metal ion transporter family 1.A.36 Intracellular chloride channel family 1.A.37 CD20 Ca2+ channel family 1.A.38 Golgi pH regulator family 1.A.39 Type C influenza virus CM2 channel family 1.A.40 Human immunodeficiency virus type I Vpu channel family 1.A.41 Avian reovirus p10 Vvroporin family 1.A.42 HIV viral protein R family 1.A.43 Camphor resistance or fluoride exporter family 1.A.44 Pore-forming tail Tip pb2 protein of phage T5 family 1.A.45 Phage P22 injectisome family 1.A.46 Anion channel-forming bestrophin family 1.A.47 Nucleotide-sensitive anion-selective channel, ICln family 1.A.48 Anion channel Tweety family 1.A.49 Human coronavirus ns12.9 viroporin family 1.A.50 Phospholamban (Ca2+-channel and Ca2+-ATPase regulator) family 1.A.51 The Voltage-gated Proton Channel (VPC) Family 1.A.52 The Ca2+ Release-activated Ca2+ (CRAC) Channel (CRAC-C) Family 1.A.53 The Hepatitis C Virus P7 Viroporin Cation-selective Channel (HCV-P7) Family 1.A.54 The Presenilin ER Ca2+ Leak Channel (Presenilin) Family 1.A.55 The Synaptic Vesicle-Associated Ca2+ Channel, Flower (Flower) Family 1.A.56 The Copper Transporter (Ctr) Family 1.A.57 The Human SARS Coronavirus Viroporin (SARS-VP) 1.A.58 The Type B Influenza Virus Matrix Protein 2 (BM2-C) Family 1.A.59 The Bursal Disease Virus Pore-Forming Peptide, Pep46 (Pep46) Family 1.A.60 The Mammalian Reovirus Pre-forming Peptide, Mu-1 (Mu-1) Family 1.A.61 The Insect Nodavirus Channel-forming Chain F (Gamma-Peptide) Family 1.A.62 The Homotrimeric Cation Channel (TRIC) Family 1.A.63 The Ignicoccus Outer Membrane α-helical Porin (I-OMP Family 1.A.64 The Plasmolipin (Plasmolipin) Family 1.A.65 The Coronavirus Viroporin E Protein (Viroporin E) Family 1.A.66 The Pardaxin (Pardaxin) Family 1.A.67 The Membrane Mg2+ Transporter (MMgT) Family 1.A.68 The Viral Small Hydrophobic Viroporin (V-SH) Family 1.A.69 The Heteromeric Odorant Receptor Channel (HORC) Family 1.A.70 The Molecule Against Microbes A (MamA) Family 1.A.71 The Brain Acid-soluble Protein Channel (BASP1 Channel) Family 1.A.72 The Mer Superfamily 1.A.73 The Colicin Lysis Protein (CLP) Family 1.A.74 The Mitsugumin 23 (MG23) Family 1.A.75 The Mechanical Nociceptor, Piezo (Piezo) Family 1.A.76 The Magnesium Transporter1 (MagT1) Family 1.A.77 The Mg2+/Ca2+ Uniporter (MCU) Family 1.A.78 The K+-selective Channel in Endosomes and Lysosomes (KEL) Family 1.A.79 The Cholesterol Uptake Protein (ChUP) or Double Stranded RNA Uptake Family 1.A.80 The NS4a Viroporin (NS4a) Family 1.A.81 The Low Affinity Ca2+ Channel (LACC) Family 1.A.82 The Hair Cell Mechanotransduction Channel (HCMC) Family 1.A.83 The SV40 Virus Viroporin VP2 (SV40 VP2) Family 1.A.84 The Calcium Homeostasis Modulator Ca2+ Channel (CALHM-C) Family 1.A.85 The Poliovirus 2B Viroporin (2B Viroporin) Family 1.A.86 The Human Papilloma Virus type 16 (HPV16) L2 Viroporin (L2 Viroporin) Family 1.A.87 The Mechanosensitive Calcium Channel (MCA) Family 1.A.88 The Fungal Potassium Channel (F-Kch) Family 1.A.89 The Human Coronavirus 229E Viroporin (229E Viroporin) Family 1.A.90 The Human Metapneumovirus (HMPV) Viroporin (HMPV-Viroporin) Family 1.A.91 The Cytoadherence-linked Asexual Protein 3.2 of Plasmodium falciparum (Clag3) Family 1.A.92 The Reovirus Viroporin VP10 (RVP10) Family 1.A.93 The Bluetongue Virus Non-Structural Protein 3 Viroporin (NS3) Family 1.A.94 The Rotavirus Non-structural Glycoprotein 4 Viroporin (NSP4) Family 1.A.95 The Ephemerovirus Viroporin (EVVP) Family 1.A.96 The Human Polyoma Virus Viroporin (PVVP) Family 1.A.97 The Human Papillomavirus type 16 E5 Viroporin (HPV-E5) Family 1.A.98 Human T-Lymphotropic Virus 1 P13 protein (HTLV1-P13) Family 1.A.99 The Infectious Bronchitis Virus Envelope Small Membrane Protein E (IBV-E) Family 1.A.100 The Rhabdoviridae Putative Viroporin, U5 (RV-U5) Family 1.A.101 The Peroxisomal Pore-forming Pex11 (Pex11) Family 1.A.102 Influenza A viroporin PB1-F2 (PB1-F2) Family 1.A.103 The Simian Virus 5 (Parainfluenza Virus 5) SH (SV5-SH) Family 1.A.104 The Proposed Flagellar Biosynthesis Na+ Channel, FlaH (FlaH) Family 1.A.105 The Mixed Lineage Kinase Domain-like (MLKL) Family 1.A.106 The Calcium Load-activated Calcium Channel (CLAC) Family 1.A.107 The Pore-forming Globin (Globin) Family
===== Process ===== A large fraction of human tumor mutations are effectively patient-specific. Therefore, neoantigens may also be based on individual tumor genomes. Deep-sequencing technologies can identify mutations within the protein-coding part of the genome (the exome) and predict potential neoantigens. In mice models, for all novel protein sequences, potential MHC-binding peptides were predicted. The resulting set of potential neoantigens was used to assess T cell reactivity. Exome–based analyses were exploited in a clinical setting, to assess reactivity in patients treated by either tumor-infiltrating lymphocyte (TIL) cell therapy or checkpoint blockade. Neoantigen identification was successful for multiple experimental model systems and human malignancies. The false-negative rate of cancer exome sequencing is low—i.e.: the majority of neoantigens occur within exonic sequence with sufficient coverage. However, the vast majority of mutations within expressed genes do not produce neoantigens that are recognized by autologous T cells. As of 2015 mass spectrometry resolution is insufficient to exclude many false positives from the pool of peptides that may be presented by MHC molecules. Instead, algorithms are used to identify the most likely candidates. These algorithms consider factors such as the likelihood of proteasomal processing, transport into the endoplasmic reticulum, affinity for the relevant MHC class I alleles and gene expression or protein translation levels.
Sources: en.wikipedia.org
Collagen is a long, triple-helical structural protein. Collagen peptides are shorter fragments made by hydrolysis, which removes the helix and improves water solubility. The two materials differ in molecular size, viscosity, and behavior in solution.
No. Chain length, amino acid profile, and trace composition vary with raw material and hydrolysis conditions. Products from fish, bovine, and porcine sources can differ in odor, color, and thermal behavior. The term covers a broad family rather than one uniform substance.
Glycine, proline, and hydroxyproline are especially abundant. Hydroxyproline is uncommon in most other proteins and is often used as a marker for collagen content. The peptides also contain varying amounts of alanine, arginine, and other residues.
Size-exclusion chromatography is the most common method, often calibrated with protein standards of known molecular weight. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) can provide a visual profile. Mass spectrometry is used for detailed peptide sequencing.