Impact of AlphaFold on structure prediction of protein complexes: The CASP15-CAPRI experiment.

We present the results for CAPRI Round 54, the 5th joint CASP-CAPRI protein assembly prediction challenge. The Round offered 37 targets, including 14 homodimers, 3 homo-trimers, 13 heterodimers including 3 antibody-antigen complexes, and 7 large assemblies. On average ~70 CASP and CAPRI predictor groups, including more than 20 automatics servers, submitted models for each target. A total of 21 941 models submitted by these groups and by 15 CAPRI scorer groups were evaluated using the CAPRI model quality measures and the DockQ score consolidating these measures. The prediction performance was quantified by a weighted score based on the number of models of acceptable quality or higher submitted by each group among their five best models. Results show substantial progress achieved across a significant fraction of the 60+ participating groups. High-quality models were produced for about 40% of the targets compared to 8% two years earlier. This remarkable improvement is due to the wide use of the AlphaFold2 and AlphaFold2-Multimer software and the confidence metrics they provide. Notably, expanded sampling of candidate solutions by manipulating these deep learning inference engines, enriching multiple sequence alignments, or integration of advanced modeling tools, enabled top performing groups to exceed the performance of a standard AlphaFold2-Multimer version used as a yard stick. This notwithstanding, performance remained poor for complexes with antibodies and nanobodies, where evolutionary relationships between the binding partners are lacking, and for complexes featuring conformational flexibility, clearly indicating that the prediction of protein complexes remains a challenging problem.

WT1 regulates expression of DNA repair gene Neil3 during nephrogenesis.

Mammalian nephrons arise from a population of nephron progenitor cells (NPCs) expressing the master transcription factor Wilms tumor-1 (WT1), which is crucial for NPC proliferation, migration, and differentiation. In humans, biallelic loss of WT1 precludes nephrogenesis and leads to the formation of Wilms tumor precursor lesions. We hypothesize that WT1 normally primes the NPC for nephrogenesis by inducing expression of NPC-specific DNA repair genes that protect the genome. We analyzed transcript levels for a panel of DNA repair genes in embryonic day 17.5 (E17.5) versus adult mouse kidneys and noted seven genes that were increased >20-fold. We then isolated Cited1+ NPCs from E17.5 kidneys and found that only one gene, nei-like DNA glycosylase 3 (Neil3), was enriched. RNAscope in situ hybridization of E17.5 mouse kidneys showed increased Neil3 expression in the nephrogenic zone versus mature nephron structures. To determine whether Neil3 expression is WT1 dependent, we knocked down Wt1 in Cited1+ NPCs (60% knockdown efficiency) and noted a 58% reduction in Neil3 transcript levels. We showed that WT1 interacts with the Neil3 promoter and that activity of a Neil3 promoter-reporter vector was increased twofold in WT1+ versus WT1- cells. We propose that Neil3 is a WT1-dependent DNA repair gene expressed at high levels in Cited1+ NPCs, where it repairs mutational injury to the genome during nephrogenesis. NEIL3 is likely just one of many such lineage-specific repair mechanisms that respond to genomic injury during kidney development.NEW & NOTEWORTHY We studied the molecular events leading to Wilms tumors as a model for the repair of genomic injury. Specifically, we showed that WT1 activates DNA repair gene Neil3 in nephron progenitor cells. However, our observations offer a much broader principle, demonstrating that the embryonic kidney invests in lineage-specific expression of DNA repair enzymes. Thus, it is conceivable that failure of these mechanisms could lead to a variety of "sporadic" congenital renal malformations and human disease.

End of induction patient reported outcomes predict clinical remission and endoscopic improvement with vedolizumab and adalimumab in ulcerative colitis.

BACKGROUND: Patient-reported outcomes (PROs) are increasingly emphasized as endpoints in clinical trials of ulcerative colitis (UC). However, the prognostic value of early improvement in PROs for long-term outcomes remains unclear. METHODS: This was a post-hoc analysis of 611 vedolizumab-treated or adalimumab-treated patients in the VARSITY trial (Clinicaltrial.gov: NCT02497469). Stool frequency (SF) and rectal bleeding score (RBS) as reported in the Mayo score at post-induction (week 6 and 14) was assessed for their association with one-year endoscopic improvement (EI), defined as Mayo endoscopic subscore <2; histo-endoscopic mucosal improvement (HEMI), defined as EI and Geboes highest grade <3.2, clinical remission (CR), defined as total Mayo score ≤2; and PRO-2 remission, defined as RBS of 0 and SF ≤1. Multivariable logistic regression models adjusted for confounders assessed the relationships between post-induction PROs and outcomes of interest at one-year. RESULTS: Patients with severe SF at week 6 were significantly less likely to achieve one-year EI compared to those with non-severe SF [aOR 0.40 (95% CI: 0.24-0.68), p < .001]. Absence of rectal bleeding at week 6 was associated with greater odds of achieving EI at one-year [aOR 2.21 (95% CI: 1.58-3.09), p < .001]. These findings were consistent across comparisons at week 14. Similar findings were observed for the outcomes of one-year HEMI, CR and PRO-2 remission. No difference was observed between the modified partial Mayo score and modified PRO-2 score. CONCLUSIONS: Post-induction PROs strongly predict the odds of CR and EI in UC and simplified evaluations can be used to assess early response to UC therapies.

Chemical Properties of Peanut Oil from Arachis hypogaea L. 'Tainan 14' and Its Oxidized Volatile Formation.

Arachis hypogaea L. 'Tainan 14' has purple skin characteristics. This study investigated the effects of different materials (shelled or unshelled peanuts) and temperatures (120 or 140 °C) on the properties of extracted peanut oil. The results show that its antioxidant components (total flavonoid, α−tocopherol, and γ-tocopherol) and oxidative stability were mainly affected by the roasting temperature (p < 0.05). Fifty-eight volatile compounds were identified by peanut oil oxidation and divided into three main groups during the roasting process using principal component analysis. The volatile formation changes of different materials and temperatures were assessed by agglomerative hierarchical clustering analysis. These results provide useful reference information for peanut oil applications in the food industry.

Analysis of Volatile Compounds from Different Parts of Houttuynia cordata Thunb.

Houttuynia cordata Thunb. is a medicinal and edible plant that has been commonly used in traditional Chinese medicine since ancient times. This study used headspace solid-phase microextraction (HS-SPME) and direct injection, combined with gas chromatography (GC) and gas chromatography-mass spectrometry (GC-MS), to identify the volatile compounds in H. cordata. Extraction from different parts of the plant using different extraction techniques for the identification of volatile compounds were determined. A total of 93 volatile components were analyzed in the leaves, stems, rhizomes, and whole plant samples of H. cordata. The leaves contained more (Z)-3-hexenal, ß-myrcene, (Z)-ß-ocimene, and (4E,6E)-allo-ocimene; the stems contained more geranyl acetate and nerolidol; and rhizomes contained more α-pinene, ß-pinene, limonene, 2-undecanone, and decanoyl acetaldehyde. Among them, the essential oil extracted by HS-SPME could produce more monoterpenes, while direct injection could obtain higher contents of aliphatic ketones, terpene esters, sesquiterpenes, and was more conducive to the extraction of 2-undecanone and decanoyl acetaldehyde.

Effects of Different Extraction Methods on Vanilla Aroma.

To establish the analytic conditions for examining the aroma quality of vanilla pods, we compared different extraction methods and identified a suitable option. We utilized headspace solid-phase microextraction (HS-SPME), steam distillation (SD), simultaneous steam distillation (SDE) and alcoholic extraction combined with gas chromatography (GC) and gas chromatography-mass spectrometry (GC-MS) to identify volatile components of vanilla pods. A total of 84 volatile compounds were identified in this experiment, of which SDE could identify the most volatile compounds, with a total of 51 species, followed by HS-SPME, with a total of 28 species. Ten volatile compounds were identified by extraction with a minimum of 35% alcohol. HS-SPME extraction provided the highest total aroma peak areas, and the peak areas of aldehydes, furans, alcohols, monoterpenes and phenols compounds were several times higher than those of the other extraction methods. The results showed that the two technologies, SDE and HS-SPME, could be used together to facilitate analysis of vanilla pod aroma.

Structural insights into nanoRNA degradation by human Rexo2.

Human RNA exoribonuclease 2 (Rexo2) is an evolutionarily conserved 3'-to-5' DEDDh-family exonuclease located primarily in mitochondria. Rexo2 degrades small RNA oligonucleotides of <5 nucleotides (nanoRNA) in a way similar to Escherichia coli Oligoribonuclease (ORN), suggesting that it plays a role in RNA turnover in mitochondria. However, how Rexo2 preferentially binds and degrades nanoRNA remains elusive. Here, we show that Rexo2 binds small RNA and DNA oligonucleotides with the highest affinity, and it is most robust in degrading small nanoRNA into mononucleotides in the presence of magnesium ions. We further determined three crystal structures of Rexo2 in complex with single-stranded RNA or DNA at resolutions of 1.8-2.2 Å. Rexo2 forms a homodimer and interacts mainly with the last two 3'-end nucleobases of substrates by hydrophobic and π-π stacking interactions via Leu53, Trp96, and Tyr164, signifying its preference in binding and degrading short oligonucleotides without sequence specificity. Crystal structure of Rexo2 is highly similar to that of the RNA-degrading enzyme ORN, revealing a two-magnesium-ion-dependent hydrolysis mechanism. This study thus provides the molecular basis for human Rexo2, showing how it binds and degrades nanoRNA into nucleoside monophosphates and plays a crucial role in RNA salvage pathways in mammalian mitochondria.

Structural basis for overhang excision and terminal unwinding of DNA duplexes by TREX1.

Three prime repair exonuclease 1 (TREX1) is an essential exonuclease in mammalian cells, and numerous in vivo and in vitro data evidenced its participation in immunity regulation and in genotoxicity remediation. In these very complicated cellular functions, the molecular mechanisms by which duplex DNA substrates are processed are mostly elusive because of the lack of structure information. Here, we report multiple crystal structures of TREX1 complexed with various substrates to provide the structure basis for overhang excision and terminal unwinding of DNA duplexes. The substrates were designed to mimic the intermediate structural DNAs involved in various repair pathways. The results showed that the Leu24-Pro25-Ser26 cluster of TREX1 served to cap the nonscissile 5'-end of the DNA for precise removal of the short 3'-overhang in L- and Y-structural DNA or to wedge into the double-stranded region for further digestion along the duplex. Biochemical assays were also conducted to demonstrate that TREX1 can indeed degrade double-stranded DNA (dsDNA) to a full extent. Overall, this study provided unprecedented knowledge at the molecular level on the enzymatic substrate processing involved in prevention of immune activation and in responses to genotoxic stresses. For example, Arg128, whose mutation in TREX1 was linked to a disease state, were shown to exhibit consistent interaction patterns with the nonscissile strand in all of the structures we solved. Such structure basis is expected to play an indispensable role in elucidating the functional activities of TREX1 at the cellular level and in vivo.

Crystal structure of dimeric human PNPase reveals why disease-linked mutants suffer from low RNA import and degradation activities.

Human polynucleotide phosphorylase (PNPase) is an evolutionarily conserved 3'-to-5' exoribonuclease principally located in mitochondria where it is responsible for RNA turnover and import. Mutations in PNPase impair structured RNA transport into mitochondria, resulting in mitochondrial dysfunction and disease. PNPase is a trimeric protein with a doughnut-shaped structure hosting a central channel for single-stranded RNA binding and degradation. Here, we show that the disease-linked human PNPase mutants, Q387R and E475G, form dimers, not trimers, and have significantly lower RNA binding and degradation activities compared to wild-type trimeric PNPase. Moreover, S1 domain-truncated PNPase binds single-stranded RNA but not the stem-loop signature motif of imported structured RNA, suggesting that the S1 domain is responsible for binding structured RNAs. We further determined the crystal structure of dimeric PNPase at a resolution of 2.8 Å and, combined with small-angle X-ray scattering, show that the RNA-binding K homology and S1 domains are relatively inaccessible in the dimeric assembly. Taken together, these results show that mutations at the interface of the trimeric PNPase tend to produce a dimeric protein with destructive RNA-binding surfaces, thus impairing both of its RNA import and degradation activities and leading to mitochondria disorders.

Structural insights into RNA unwinding and degradation by RNase R.

RNase R is a conserved exoribonuclease in the RNase II family that primarily participates in RNA decay in all kingdoms of life. RNase R degrades duplex RNA with a 3' overhang, suggesting that it has RNA unwinding activity in addition to its 3'-to-5' exoribonuclease activity. However, how RNase R coordinates RNA binding with unwinding to degrade RNA remains elusive. Here, we report the crystal structure of a truncated form of Escherichia coli RNase R (residues 87-725) at a resolution of 1.85 Å. Structural comparisons with other RNase II family proteins reveal two open RNA-binding channels in RNase R and suggest a tri-helix 'wedge' region in the RNB domain that may induce RNA unwinding. We constructed two tri-helix wedge mutants and they indeed lost their RNA unwinding but not RNA binding or degrading activities. Our results suggest that the duplex RNA with an overhang is bound in the two RNA-binding channels in RNase R. The 3' overhang is threaded into the active site and the duplex RNA is unwound upon reaching the wedge region during RNA degradation. Thus, RNase R is a proficient enzyme, capable of concurrently binding, unwinding and degrading structured RNA in a highly processive manner during RNA decay.

Selective influence of Sox2 on POU transcription factor binding in embryonic and neural stem cells.

Embryonic stem cell (ESC) identity is orchestrated by co-operativity between the transcription factors (TFs) Sox2 and the class V POU-TF Oct4 at composite Sox/Oct motifs. Neural stem cells (NSCs) lack Oct4 but express Sox2 and class III POU-TFs Oct6, Brn1 and Brn2. This raises the question of how Sox2 interacts with POU-TFs to transcriptionally specify ESCs versus NSCs. Here, we show that Oct4 alone binds the Sox/Oct motif and the octamer-containing palindromic MORE equally well. Sox2 binding selectively increases the affinity of Oct4 for the Sox/Oct motif. In contrast, Oct6 binds preferentially to MORE and is unaffected by Sox2. ChIP-Seq in NSCs shows the MORE to be the most enriched motif for class III POU-TFs, including MORE subtypes, and that the Sox/Oct motif is not enriched. These results suggest that in NSCs, co-operativity between Sox2 and class III POU-TFs may not occur and that POU-TF-driven transcription uses predominantly the MORE cis architecture. Thus, distinct interactions between Sox2 and POU-TF subclasses distinguish pluripotent ESCs from multipotent NSCs, providing molecular insight into how Oct4 alone can convert NSCs to pluripotency.

Wilms tumor suppressor, WT1, suppresses epigenetic silencing of the ß-catenin gene.

The mammalian kidney is derived from progenitor cells in intermediate mesoderm. During embryogenesis, progenitor cells expressing the Wilms tumor suppressor gene, WT1, are induced to differentiate in response to WNT signals from the ureteric bud. In hereditary Wilms tumors, clonal loss of WT1 precludes the ß-catenin pathway response and leads to precancerous nephrogenic rests. We hypothesized that WT1 normally primes progenitor cells for differentiation by suppressing the enhancer of zeste2 gene (EZH2), involved in epigenetic silencing of differentiation genes. In human amniotic fluid-derived mesenchymal stem cells, we show that exogenous WT1B represses EZH2 transcription. This leads to a dramatic decrease in the repressive lysine 27 trimethylation mark on histone H3 that silences ß-catenin gene expression. As a result, amniotic fluid mesenchymal stem cells acquire responsiveness to WNT9b and increase expression of genes that mark the onset of nephron differentiation. Our observations suggest that biallelic loss of WT1 sustains the inhibitory histone methylation state that characterizes Wilms tumors.

Cognitive and epilepsy outcomes after epilepsy surgery caused by focal cortical dysplasia in children: early intervention maybe better.

BACKGROUND: Focal cortical dysplasia (FCD) is a specific malformation of cortical development harboring intrinsic epileptogenicity, and most of the patients develop drug-resistant epilepsy in early childhood. The detrimental effects of early and frequent seizures on cognitive function in children are significant clinical issues. In this study, we evaluate the effects of early surgical intervention of FCD on epilepsy outcome and cognitive development. METHODS: From 2006 to 2013, 30 children younger than 18 years old underwent resective surgery for FCDs at Taipei Veterans General Hospital. The mean age at surgery was 10.0 years (range 1.7 to 17.6 years). There were 21 boys and 9 girls. In this retrospective clinical study, seizure outcome, cognitive function, and quality of life were evaluated. To evaluate the effects to outcomes on early interventions, the patients were categorized into four groups according to age of seizure onset, duration of seizure before surgery, and severity of cognitive deficits. RESULTS: Eleven of 22 (50 %) patients demonstrated developmental delay preoperatively. The Engel seizure outcome achievements were class I in 21 (70 %), class II in 2 (7 %), class III in 6 (20 %), and class IV in 1 (3 %) patients. The locations of FCDs resected were in the frontal lobe in 18 cases, temporal lobe in 7, parietal lobe in 2, and in bilobes including frontoparietal lobe in 2 and parieto-occipital lobes in 1. Eight cases that had FCDs involved in the rolandic cortex presented hemiparesis before surgical resection. Motor function in four of them improved after operation. The histopathological types of FCDs were type Ia in 1, type Ib in 7, type IIa in 7, type IIb in 12, and type III in 3 patients. FCDs were completely resected in 20 patients. Eighteen (90 %) of them were seizure free (p < 0.001) with three patients that received more than one surgery to accomplish complete resection. The patients who had early seizure onset, no significant cognitive function deficit, and early surgical intervention with complete resection in less than 2 years of seizure duration showed best outcomes on seizure control, cognitive function, and quality of life. CONCLUSION: Delay in cognitive development and poor quality of life is common in children treated for FCDs. Early surgical intervention and complete resection of the lesion help for a better seizure control, cognitive function development, and quality of life. FCDs involved eloquent cortex may not prohibit complete resection for better outcomes.

Flexible protein-protein docking with a multitrack iterative transformer.

Conventional protein-protein docking algorithms usually rely on heavy candidate sampling and reranking, but these steps are time-consuming and hinder applications that require high-throughput complex structure prediction, for example, structure-based virtual screening. Existing deep learning methods for protein-protein docking, despite being much faster, suffer from low docking success rates. In addition, they simplify the problem to assume no conformational changes within any protein upon binding (rigid docking). This assumption precludes applications when binding-induced conformational changes play a role, such as allosteric inhibition or docking from uncertain unbound model structures. To address these limitations, we present GeoDock, a multitrack iterative transformer network to predict a docked structure from separate docking partners. Unlike deep learning models for protein structure prediction that input multiple sequence alignments, GeoDock inputs just the sequences and structures of the docking partners, which suits the tasks when the individual structures are given. GeoDock is flexible at the protein residue level, allowing the prediction of conformational changes upon binding. On the Database of Interacting Protein Structures (DIPS) test set, GeoDock achieves a 43% top-1 success rate, outperforming all other tested methods. However, in the standard DIPS train/test splits, we discovered contamination of close homologs in the training set. After decontaminating the training set, the success rate is 31%. On the DB5.5 test set and a benchmark dataset of antibody-antigen complexes, GeoDock outperforms the deep learning models trained using the same dataset but falls behind most of the conventional methods and AlphaFold-Multimer. GeoDock attains an average inference speed of under 1 s on a single GPU, enabling its application in large-scale structure screening. Although binding-induced conformational changes are still a challenge owing to limited training and evaluation data, our architecture sets up the foundation to capture this backbone flexibility. Code and a demonstration Jupyter notebook are available at https://github.com/Graylab/GeoDock.

Molecular Display of the Animal Meta-Venome for Discovery of Novel Therapeutic Peptides.

Animal venoms, distinguished by their unique structural features and potent bioactivities, represent a vast and relatively untapped reservoir of therapeutic molecules. However, limitations associated with extracting or expressing large numbers of individual venoms and venom-like molecules have precluded their therapeutic evaluation via high throughput screening. Here, we developed an innovative computational approach to design a highly diverse library of animal venoms and "metavenoms". We employed programmable M13 hyperphage display to preserve critical disulfide-bonded structures for highly parallelized single-round biopanning with quantitation via high-throughput DNA sequencing. Our approach led to the discovery of Kunitz type domain containing proteins that target the human itch receptor Mas-related G protein-coupled receptor X4 (MRGPRX4), which plays a crucial role in itch perception. Deep learning-based structural homology mining identified two endogenous human homologs, tissue factor pathway inhibitor (TFPI) and serine peptidase inhibitor, Kunitz type 2 (SPINT2), which exhibit agonist-dependent potentiation of MRGPRX4. Highly multiplexed screening of animal venoms and metavenoms is therefore a promising approach to uncover new drug candidates.

Single cohesin molecules generate force by two distinct mechanisms.

Spatial organization of DNA is facilitated by cohesin protein complexes that move on DNA and extrude DNA loops. How cohesin works mechanistically as a molecular machine is poorly understood. Here, we measure mechanical forces generated by conformational changes in single cohesin molecules. We show that bending of SMC coiled coils is driven by random thermal fluctuations leading to a ~32 nm head-hinge displacement that resists forces up to 1 pN; ATPase head engagement occurs in a single step of ~10 nm and is driven by an ATP dependent head-head movement, resisting forces up to 15 pN. Our molecular dynamic simulations show that the energy of head engagement can be stored in a mechanically strained conformation of NIPBL and released during disengagement. These findings reveal how single cohesin molecules generate force by two distinct mechanisms. We present a model, which proposes how this ability may power different aspects of cohesin-DNA interaction.

Flexible Protein-Protein Docking with a Multi-Track Iterative Transformer.

Conventional protein-protein docking algorithms usually rely on heavy candidate sampling and re-ranking, but these steps are time-consuming and hinder applications that require high-throughput complex structure prediction, e.g., structure-based virtual screening. Existing deep learning methods for protein-protein docking, despite being much faster, suffer from low docking success rates. In addition, they simplify the problem to assume no conformational changes within any protein upon binding (rigid docking). This assumption precludes applications when binding-induced conformational changes play a role, such as allosteric inhibition or docking from uncertain unbound model structures. To address these limitations, we present GeoDock, a multi-track iterative transformer network to predict a docked structure from separate docking partners. Unlike deep learning models for protein structure prediction that input multiple sequence alignments (MSAs), GeoDock inputs just the sequences and structures of the docking partners, which suits the tasks when the individual structures are given. GeoDock is flexible at the protein residue level, allowing the prediction of conformational changes upon binding. For a benchmark set of rigid targets, GeoDock obtains a 41% success rate, outperforming all the other tested methods. For a more challenging benchmark set of flexible targets, GeoDock achieves a similar number of top-model successes as the traditional method ClusPro [1], but fewer than ReplicaDock2 [2]. GeoDock attains an average inference speed of under one second on a single GPU, enabling its application in large-scale structure screening. Although binding-induced conformational changes are still a challenge owing to limited training and evaluation data, our architecture sets up the foundation to capture this backbone flexibility. Code and a demonstration Jupyter notebook are available at https://github.com/Graylab/GeoDock.

Fast, accurate antibody structure prediction from deep learning on massive set of natural antibodies.

Antibodies have the capacity to bind a diverse set of antigens, and they have become critical therapeutics and diagnostic molecules. The binding of antibodies is facilitated by a set of six hypervariable loops that are diversified through genetic recombination and mutation. Even with recent advances, accurate structural prediction of these loops remains a challenge. Here, we present IgFold, a fast deep learning method for antibody structure prediction. IgFold consists of a pre-trained language model trained on 558 million natural antibody sequences followed by graph networks that directly predict backbone atom coordinates. IgFold predicts structures of similar or better quality than alternative methods (including AlphaFold) in significantly less time (under 25 s). Accurate structure prediction on this timescale makes possible avenues of investigation that were previously infeasible. As a demonstration of IgFold's capabilities, we predicted structures for 1.4 million paired antibody sequences, providing structural insights to 500-fold more antibodies than have experimentally determined structures.

Preliminary study on the application of bioimpedance analysis to measure the psoas major muscle in older adults.

For the assessment of sarcopenia or other geriatric frailty syndromes, psoas major area may be one of the primary indicators. Aim to develop and cross-validate the psoas cross-sectional area estimation equation of L3-L4 of the elderly over 60 years old by bioelectrical impedance analysis (BIA). Ninety-two older adults with normal mobility were enrolled (47 females, 45 males), and were randomly divided into a modeling group (MG, n = 62) and validation group (VG, n = 30). Computed tomography (CT) was used to measure the psoas major area at the' L3-L4 lumbar vertebrae height as a predictor. Estimated variables were height (h), whole body impedance (Zwhole), whole body impedance index (h2/Zwhole, WBI), age, gender (female = 0, male = 1), and body weight (weight) by standing BIA. Relevant variables were estimated using stepwise regression analysis. Model performance was confirmed by cross-validation. BIA estimation equation for PMM obtained from the MG was: (PMMBIA = 0.183 h2/Z- 0.223 age + 4.443 gender + 5.727, r2 = 0.702, n = 62, SEE = 2.432 cm2, p < 0.001). The correlation coefficient r obtained by incorporating the VG data into the PMM equation was 0.846, and the LOA ranged from -4.55 to 4.75 cm2. PMMBIA and PMMCT both correlate highly with MG or VG with small LOA. The fast and convenient standing BIA for measuring PMM may be a promising method that is worth developing.

Discussion on the Differences in Aroma Components in Different Fragrant Rice Varieties during Storage.

Rice (Oryza sativa L.) is an important food crop in Taiwan, among which fragrant rice is highly regarded for its special aroma when cooked. During the storage of fragrant rice, the aroma components will change, which will affect the aroma quality of fragrant rice. Therefore, headspace solid-phase microextraction (HS-SPME) was used in this study, combined with gas chromatography and gas chromatography-mass spectrometry, to analyze the difference in the aroma components of Taikeng No. 4 (TK4), Tainung No. 71 (TN71), Kaohsiung No. 147 (KH147), and Taichung No. 194 (TC194) fragrant rice. A total of 28 aroma components were identified in the four varieties of fragrant rice, and the main components were all Nonanal. Among them, TK4 contains a very high content of hydrocarbons, including Tridecane and Dodecane; TN71, KH147, and TC194 contain mainly aldehydes such as Nonanal and Hexanal. During different storage times, the contents of alcohols, monoterpenes, aromatic aldehydes, and furans increased with storage time, while the content of aliphatic aldehydes decreased with storage time. After storage, the fragrant rice samples showed a tendency for the total volatile component content to decrease, with the most pronounced reduction observed in Nonanal content.