RESUMEN
Conventional federated learning (FL) assumes the homogeneity of models, necessitating clients to expose their model parameters to enhance the performance of the server model. However, this assumption cannot reflect real-world scenarios. Sharing models and parameters raises security concerns for users, and solely focusing on the server-side model neglects clients' personalization requirements, potentially impeding expected performance improvements of users. On the other hand, prioritizing personalization may compromise the generalization of the server model, thereby hindering extensive knowledge migration. To address these challenges, we put forth an important problem: How can FL ensure both generalization and personalization when clients' models are heterogeneous? In this work, we introduce FedTED, which leverages a twin-branch structure and data-free knowledge distillation (DFKD) to address the challenges posed by model heterogeneity and diverse objectives in FL. The employed techniques in FedTED yield significant improvements in both personalization and generalization, while effectively coordinating the updating process of clients' heterogeneous models and successfully reconstructing a satisfactory global model. Our empirical evaluation demonstrates that FedTED outperforms many representative algorithms, particularly in scenarios where clients' models are heterogeneous, achieving a remarkable 19.37% enhancement in generalization performance and up to 9.76% improvement in personalization performance.
RESUMEN
The Chinese goral (Naemorhedus griseus) is identified as a vulnerable species on the Red List of China's Biodiversity and listed as a national second-class key protected wild animal in China. It is a representative flagship ungulate in Beijing. Its distribution range is fragmented and small populations are separated by dense infrastructures and tourism sites. Understanding its population status provides a foundation to plan effective conservation strategies. In this study, a population viability analysis was conducted with VORTEX (10.5.6.0) on a Chinese goral population in Beijing Yunmengshan Nature Reserve with the data collected by camera trapping and parameters referenced from other goral populations. The results show that this population will initially increase in the next 20 years and then decrease with a 32% probability of extinction risk. Supplementation with four adults, two females and two males, every 10 years would help minimize the extinction risk of this population. The results highlight the vital limiting factors for Chinese goral, including the initial population size, sex ratio at birth and mortality of infants (especially female infants). To improve the protection efficiency, detailed population parameters should be further acquired through continuous monitoring of this population. A thorough large-scale study should be carried out on other segregated goral populations in Beijing to facilitate the recovery of this endangered species.
RESUMEN
During the COVID-19 pandemic, numerous people experiencing illness or senescence choose to receive home health care (HHC) services. However, a rapid increase in patients makes it a challenge to reasonably allocate nurses to provide HHC services under the condition of a paucity of nurse resources and patient time window constraints. To solve the large-scale HHC problem, a hybrid heuristic-exact optimization algorithm is proposed with three novel contributions. Firstly, a framework of hybrid heuristic-exact optimization is designed to solve the large-scale problem where a reasonable solution is difficult to obtain under constraints. Secondly, a multi-objective mixed-integer linear programming modelization is formulated to get a more diverse nurse assignment. Finally, an improved branch and bound algorithm is proposed to speed up computation for the large-scale problem. Computational results on different HHC instances from 25 to 1000 patients demonstrate that the proposed algorithm can optimize the HHC problem with more than 100 patients and can provide various assignments for different numbers of nurses, which the common algorithm cannot optimize.
RESUMEN
Programmable RNA-guided DNA nucleases perform numerous roles in prokaryotes, but the extent of their spread outside prokaryotes is unclear. Fanzors, the eukaryotic homolog of prokaryotic TnpB proteins, have been detected in genomes of eukaryotes and large viruses, but their activity and functions in eukaryotes remain unknown. Here, we characterize Fanzors as RNA-programmable DNA endonucleases, using biochemical and cellular evidence. We found diverse Fanzors that frequently associate with various eukaryotic transposases. Reconstruction of Fanzors evolution revealed multiple radiations of RuvC-containing TnpB homologs in eukaryotes. Fanzor genes captured introns and proteins acquired nuclear localization signals, indicating extensive, long-term adaptation to functioning in eukaryotic cells. Fanzor nucleases contain a rearranged catalytic site of the RuvC domain, similar to a distinct subset of TnpBs, and lack collateral cleavage activity. We demonstrate that Fanzors can be harnessed for genome editing in human cells, highlighting the potential of these widespread eukaryotic RNA-guided nucleases for biotechnology applications.
Asunto(s)
Eucariontes , Virus , Humanos , Eucariontes/genética , Eucariontes/metabolismo , Desoxirribonucleasa I , ARN/genética , Desoxirribonucleasas/metabolismo , Virus/genéticaRESUMEN
DDD/E transposase gene is the most abundant gene in nature and many DNA transposons in all three domains of life use it for their transposition. A substantial number of eukaryotic DNA transposons show similarity to prokaryotic insertion sequences (ISs). The presence of IS481-like DNA transposons was indicated in the genome of Trichomonas vaginalis. Here, we surveyed IS481-like eukaryotic sequences using a bioinformatics approach and report a group of eukaryotic IS481-like DNA transposons, designated IS481EU, from parabasalids including T. vaginalis. The lengths of target site duplications (TSDs) of IS481EU are around 4 bps, around 15 bps, or around 25 bps, and strikingly, these discrete lengths of TSDs can be observed even in a single IS481EU family. Phylogenetic analysis indicated the close relationships of IS481EU with some of the prokaryotic IS481 family members. IS481EU was not well separated from IS3EU/GingerRoot in the phylogenetic analysis, but was distinct from other eukaryotic DNA transposons including Ginger1 and Ginger2. The unique characteristics of IS481EU in protein sequences and the distribution of TSD lengths support its placement as a new superfamily of eukaryotic DNA transposons.
RESUMEN
BACKGROUND: Malignant tumors of the ileocecal region often cause intestinal obstruction. Emergency surgery is the main treatment for patients presenting with an obstruction. However, this procedure is associated with a high mortality rate and frequent complications. The placement of colon stents is commonly performed for obstructions in the distal colon and is a less invasive and safer procedure. However, obstructions in the proximal colon are more challenging to treat by stent placement due to the increased distance from the anus. CASE SUMMARY: This case report concerns an 88-year-old man with malignant intestinal obstruction in the ileocecal region. He was contraindicated for general anesthesia and surgical enterostomy. The placement of a self-expandable metallic stent seems an alternative to surgery, although stenting in this area is thought to be difficult and few studies have been reported so far. After three attempts at different interventional approaches, a stent was successfully placed in the obstructed segment under fluoroscopic guidance. After the procedure, the patient's abdominal distension and abdominal pain were significantly better than before. CONCLUSION: For patients with proximal colonic obstruction, self-expandable metallic stent placement under fluoroscopic guidance could be considered as a feasible treatment to relieve abdominal distension and pain in patients with acute bowel obstruction. It has the characteristics of high safety and high patient tolerance. However, further study is still needed.
RESUMEN
In the face of habitat shrinkage and segregation, the survival of wild cats looks bleak. Interpreting their population genetic structure during habitat fragmentation is critical in planning effective management strategies. To reveal the segregation effects of road construction and human settlements on the population genetic structure, we analyzed non-invasive fecal DNA samples from leopard cats (Prionailurus bengalensis) from five nature reserves in mountainous areas around Beijing. We focused on microsatellite markers. A total of 112 individual leopard cats were identified among 601 samples of scat, and moderate population genetic diversity was detected. Microsatellite-marker-based genetic differentiation (Fst) and gene flow (Nm) showed a weak trend toward discrepancies in the Baihuashan and Songshan subpopulations, which indicated habitat fragmentation effects on individual dispersal. Because the segregated subpopulations may suffer a high risk of genetic diversity loss, we suggest that their genetic structure be monitored with more molecular markers to detect any changes, and that female individuals be artificially introduced as needed to maintain the viability of the leopard cat subpopulations in Beijing.
RESUMEN
BACKGROUND: DNA transposons are ubiquitous components of eukaryotic genomes. A major group of them encode a DDD/E transposase and contain terminal inverted repeats (TIRs) of varying lengths. The Kolobok superfamily of DNA transposons has been found in a wide spectrum of organisms. RESULTS: Here we report a new Kolobok lineage, designated KolobokP. They were identified in 7 animal phyla (Mollusca, Phoronida, Annelida, Nemertea, Bryozoa, Chordata, and Echinodermata), and are especially rich in bivalves. Unlike other Kolobok families, KolobokP adopts a composite-like architecture: an internal region (INT) flanked by two long terminal direct repeats (LTDRs), which exhibit their own short terminal inverted repeats ranging up to 18 bps. The excision of LTDRs was strongly suggested. The LTDR lengths seem to be constrained to be either around 450-bp or around 660-bp. The internal region encodes a DDD/E transposase and a small His-Me finger nuclease, which likely originated from the homing endonuclease encoded by a group I intron from a eukaryotic species. The architecture of KolobokP resembles composite DNA transposons, usually observed in bacterial genomes, and long terminal repeat (LTR) retrotransposons. In addition to this monomeric LTDR-INT-LTDR structure, plenty of solo LTDRs and multimers represented as (LTDR-INT)n-LTDR are also observed. Our structural and phylogenetic analysis supported the birth of KolobokP in the late stage of the Kolobok evolution. We propose KolobokP families propagate themselves in two ways: the canonical transposition catalyzed by their transposase and the sequence-specific cleavage by their endonuclease followed by the multimerization through the unequal crossover. CONCLUSIONS: The presence of homing endonuclease and long terminal direct repeats of KolobokP families suggest their unique dual replication mechanisms: transposition and induced unequal crossover.
RESUMEN
BACKGROUND: Clinical notes record the health status, clinical manifestations and other detailed information of each patient. The International Classification of Diseases (ICD) codes are important labels for electronic health records. Automatic medical codes assignment to clinical notes through the deep learning model can not only improve work efficiency and accelerate the development of medical informatization but also facilitate the resolution of many issues related to medical insurance. Recently, neural network-based methods have been proposed for the automatic medical code assignment. However, in the medical field, clinical notes are usually long documents and contain many complex sentences, most of the current methods cannot effective in learning the representation of potential features from document text. METHODS: In this paper, we propose a hybrid capsule network model. Specifically, we use bi-directional LSTM (Bi-LSTM) with forwarding and backward directions to merge the information from both sides of the sequence. The label embedding framework embeds the text and labels together to leverage the label information. We then use a dynamic routing algorithm in the capsule network to extract valuable features for medical code prediction task. RESULTS: We applied our model to the task of automatic medical codes assignment to clinical notes and conducted a series of experiments based on MIMIC-III data. The experimental results show that our method achieves a micro F1-score of 67.5% on MIMIC-III dataset, which outperforms the other state-of-the-art methods. CONCLUSIONS: The proposed model employed the dynamic routing algorithm and label embedding framework can effectively capture the important features across sentences. Both Capsule networks and domain knowledge are helpful for medical code prediction task.
Asunto(s)
Clasificación Internacional de Enfermedades , Redes Neurales de la Computación , Algoritmos , Registros Electrónicos de Salud , Humanos , LenguajeRESUMEN
Detailed molecular genetic research on amphibian populations has a significant role in understanding the genetic adaptability to local environments. The oriental fire-bellied toads (Bombina orientalis) were artificially introduced to Beijing from Shandong Province in 1927, and since then, this separated population went through an independent evolution. To explore the differentiation of the introduced population with its original population, this study analyzed the genetic structure of the oriental fire-bellied toad, based on the mitochondrial genome control region and six microsatellite sites. The results showed that the haplotype diversity and nucleotide diversity of the mitochondrial D-loop region partial sequences of the Beijing Botanical Garden population and the Baiwangshan population were lower than those of the Shangdong Kunyushan population. Microsatellite marker analysis also showed that the observed heterozygosity and expected heterozygosity of the Beijing populations were lower than those of the Kunyushan population. The phylogenetic trees and network diagrams of haplotypes indicated that the three populations were not genetically separated. However, the structure analysis showed a genetic differentiation and categorized the sampling individuals into Beijing and Shandong genetic clusters, which indicated a tendency for isolated evolution in the Beijing population. Although the Beijing populations showed a decline in genetic diversity, it was still at a moderate level, which could maintain the survival of the population. Thus, there is no need to reintroduce new individuals from the Kunyushan source population.
RESUMEN
PURPOSE: To investigate risk factors for instrumentation failure (IF) in titanium (Ti) mesh reconstruction for thoracic and lumbar tumors. PATIENTS AND METHODS: The clinical data of patients with thoracic or lumbar tumors who received Ti mesh reconstruction via the posterior approach in our hospital from 2013 to 2018 were analyzed retrospectively. The observation indexes included sex, age, BMI, the vertebra resection mode, the number of resected vertebral segments, application of bone cement, radiotherapy, chemotherapy, revision or primary surgery, and primary tumor metastasis. Correlations between these factors and IF were analyzed by Kaplan-Meier survival and logistics regression analyses. RESULTS: The 178 patients included 108 males and 70 females with a mean age of 48.09±16.21 (6-78) years and a mean follow-up period of 51.18 (24-90) months. The data showed that 17 patients (9.55%) were inflicted with IF, involving the thoracic vertebra in 11 cases, thoracolumbar vertebrae (T12-L1) in 2 cases, and lumbar vertebrae in 4 cases. The mean interval between surgery to IF was 35.18±14.17 (14-59) months. Univariate analysis showed that total vertebral body resection, the number of resected vertebral segments, radiotherapy and multiple tumor resection were potential factors for IF, while multivariate analysis showed that only total vertebral body resection, the number of resected vertebral segments and radiotherapy were independent factors. CONCLUSION: Total vertebra resection, the number of resected vertebral segments (≥2) and radiotherapy before and after operation were significant risk factors related to IF.
RESUMEN
The purpose of the present study was to determine the presence of Rickettsia sp. and Coxiella burnetii in Hyalomma asiaticum collected from Bactrian camels (Camelus bactrianus) in Inner Mongolia of China. A total of 385 H. asiaticum adults collected from Bactrian camels from Alxa in Inner Mongolia from 2017 to 2018 were examined using PCR combined with sequencing. The results indicated that 18 (4.7 %) and 5 (1.3 %) samples tested positive for Rickettsia and C. burnetii, respectively. The corresponding nucleotide identities among detected genes ranged between 99.7 % and 100 % for the sequences of 17 kD and ompB from Rickettsia and 16S rRNA and FtsX from C. burnetii. Two representative strains-Rickettsia IMA-1 strain and C. burnetii IMA-1 strain-were used for subsequent analysis. NCBI BLAST and phylogenetic analyses demonstrated that the detected strain Rickettsia IMA-1 may represent a novel species of spotted fever group (SFG) rickettsiae. The C. burnetii IMA-1 strain was grouped with Coxiella burnetii str. RSA439 (GenBank: CP040059.1). In addition, the Rickettsia sp. was successfully isolated from the ticks in Vero cells incubated at 28 °C. These findings indicate that the H. asiaticum collected off Bactrian camels in Inner Mongolia carried SFG Rickettsia species and C. burnetii and could contribute to the etiology of febrile illness in animals and humans.
Asunto(s)
Camelus , Coxiella burnetii/genética , Fiebre Q/veterinaria , Infecciones por Rickettsia/veterinaria , Rickettsia/genética , Animales , China/epidemiología , Coxiella burnetii/clasificación , Genes Bacterianos , Ixodidae/microbiología , Epidemiología Molecular , Filogenia , Prevalencia , Fiebre Q/epidemiología , Fiebre Q/microbiología , Rickettsia/clasificación , Infecciones por Rickettsia/epidemiología , Infecciones por Rickettsia/microbiologíaRESUMEN
Bioinformatic and molecular characterization of satellite repeats was performed to understand the impact of their diversification on Vaccinium genome evolution. Satellite repeat diversity was evaluated in four cultivated and wild species, including the diploid species Vaccinium myrtillus and Vaccinium uliginosum, as well as the tetraploid species Vaccinium corymbosum and Vaccinium arctostaphylos. We comparatively characterized six satellite repeat families using in total 76 clones with 180 monomers. We observed that the monomer units of VaccSat1, VaccSat2, VaccSat5, and VaccSat6 showed a higher order repeat (HOR) structure, likely originating from the organization of two adjacent subunits with differing similarity, length and size. Moreover, VaccSat1, VaccSat3, VaccSat6, and VaccSat7 were found to have sequence similarity to parts of transposable elements. We detected satellite-typical tandem organization for VaccSat1 and VaccSat2 in long arrays, while VaccSat5 and VaccSat6 distributed in multiple sites over all chromosomes of tetraploid V. corymbosum, presumably in long arrays. In contrast, very short arrays of VaccSat3 and VaccSat7 are dispersedly distributed over all chromosomes in the same species, likely as internal parts of transposable elements. We provide a comprehensive overview on satellite species specificity in Vaccinium, which are potentially useful as molecular markers to address the taxonomic complexity of the genus, and provide information for genome studies of this genus.
Asunto(s)
ADN Satélite/genética , Vaccinium/genética , Cromosomas de las Plantas/genética , Biología Computacional , Elementos Transponibles de ADN , Genoma de Planta , Genotipo , Filogenia , Ploidias , Alineación de Secuencia , Especificidad de la EspecieRESUMEN
A novel hierarchical composite consisting of an ordered NiCo2O4 nanowire array growing on the lateral side of a Fe2O3 microdisc is presented, which was confirmed by X-ray holography technology on a synchrotron radiation station. The composite-based Li-ion battery anode exhibits a high capacity of 1528 mA h g-1 after 200 cycles at 0.2C, a recoverable rate-performance after repeated tests, and robust mechanical properties.
RESUMEN
White spot syndrome virus (WSSV), the lone virus of the genus Whispovirus under the family Nimaviridae, is one of the most devastating viruses affecting the shrimp farming industry. Knowledge about this virus, in particular, its evolution history, has been limited, partly due to its large genome and the lack of other closely related free-living viruses for comparative studies. In this study, we reconstructed a full-length endogenous nimavirus consensus genome, Nimav-1_LVa (279,905 bp), in the genome sequence of Penaeus (Litopenaeus) vannamei breed Kehai No. 1 (ASM378908v1). This endogenous virus seemed to insert exclusively into the telomeric pentanucleotide microsatellite (TAACC/GGTTA)n. It encoded 117 putative genes, with some containing introns, such as g012 (inhibitor of apoptosis, IAP), g046 (crustacean hyperglycemic hormone, CHH), g155 (innexin), g158 (Bax inhibitor 1 like). More than a dozen Nimav-1_LVa genes are involved in the pathogen-host interactions. We hypothesized that g046, g155, g158, and g227 (semaphorin 1A like) were recruited host genes for their roles in immune regulation. Sequence analysis indicated that a total of 43 WSSV genes belonged to the ancestral/core nimavirus gene set, including four genes reported in this study: wsv112 (dUTPase), wsv206, wsv226, and wsv308 (nucleocapsid protein). The availability of the Nimav-1_LVa sequence would help understand the genetic diversity, epidemiology, evolution, and virulence of WSSV.
Asunto(s)
Genoma Viral , Interacciones Huésped-Patógeno/genética , Nimaviridae/fisiología , Penaeidae/virología , Proteínas Virales/genética , AnimalesRESUMEN
Revealing the behavioral relationships between predators and their prey is fundamental in understanding the community structure and ecosystem functions of such animals. This study aimed at detecting the population size and activity patterns of Eurasian lynx (Lynx lynx) (along with its prey) by camera trapping monitoring from 2014 to 2017 at the Saihanwula nature reserve in central Inner Mongolia. The total effective trapping days were 29,892 and 20 lynx were identified from 343 trapping photos based on the inner side patterns of their forelimbs. The daily activity rhythms of the lynx overlapped with those of different prey in different seasons. The yearly activity pattern of the lynx was influenced by its main prey's biology. In conclusion, this study reveals that the activity patterns of the top predator matched those of its prey in different time periods. Habitat management strategies promoting the restoration of prey communities would benefit the lynx in maintaining a stable community structure.
RESUMEN
Adiponectin (ADIPOQ, encoded by Adipoq) is an important white adipose-derived adipokine linked to energy homeostasis and reproductive function. This study aims to reveal the expression and role of the adiponectin system in the ovaries under acute malnutrition. In this study, 48-h food deprivation significantly inhibited ovarian growth by suppressing cell proliferation and inducing cell apoptosis in the ovaries of gonadotrophin-primed immature mice. It was also accompanied by significantly decelerated basic metabolism (glucose, triacylglycerol and cholesterol), varied steroid hormones (follicle-stimulating hormone, luteinizing hormone and estradiol) and vanishment of the peri-ovarian fat. It is noteworthy that after acute fasting, the adiponectin levels in ovaries rather than in blood were significantly elevated. Immunohistochemical study demonstrated that adiponectin and its receptors (ADIPOR1 and ADIPOR2) primarily appeared in ovarian somatic and/or germ cells, and their protein expressions were upregulated in the ovaries from fasted mice. Further in vitro study verified that ADIPOR1/2 agonist obviously inhibited follicle-stimulating hormone-induced oocyte meiotic resumption, while the antagonist significantly enhanced the percentage of oocyte maturation in the absence of follicle-stimulating hormone. Furthermore, the build up of peri-ovarian fat under physiological status in mice showed a positive correlation with both the hypertrophy of adipocytes and growth of ovaries. Taken together, these findings indicate that the upregulation of the adiponectin system disturbs the normal female reproductive function under the malnutrition status, and it may be associated with the loss of peri-ovarian fat depots.
Asunto(s)
Adiponectina/fisiología , Restricción Calórica/efectos adversos , Ayuno/fisiología , Desnutrición/fisiopatología , Ovario/fisiología , Enfermedad Aguda , Adipocitos/fisiología , Adiponectina/farmacología , Tejido Adiposo/metabolismo , Adiposidad/fisiología , Animales , Metabolismo Energético/fisiología , Femenino , Desnutrición/complicaciones , Desnutrición/metabolismo , Desnutrición/patología , Ratones , Ovario/metabolismo , Ovario/patología , Receptores de Adiponectina/genética , Receptores de Adiponectina/metabolismo , Transducción de Señal/efectos de los fármacos , Transducción de Señal/fisiologíaRESUMEN
Detailed information on the size and genetic structure of wildlife populations is critical for developing effective conservation strategies, especially for those species that have suffered population decline and fragmentation due to anthropogenic activities. In the present study, we used a non-invasive approach combining fecal pellet sampling with mitochondrial DNA and nuclear DNA microsatellite marker analysis to monitor and compare the population structure of the Chinese goral (Naemorhedus griseus) in Beijing and northeast Inner Mongolia in China. Of the 307 fecal samples confirmed to be from N. griseus, 15 individuals (nine females and six males) were found in the Beijing population and 61 individuals (37 females and 24 males) were found in the Inner Mongolian population. Among these 76 individuals, we identified eight haplotypes and 13 nucleotide polymorphic sites from mtDNA and 45 alleles from 10 microsatellite loci. Spatially structured genetic variation and a significant level of genetic differentiation were observed between the two populations. In both populations, the sex ratios were skewed toward females, indicating high reproductive potential, which is crucial for population recovery and conservation of this patchily distributed vulnerable species. We suggest that managing the two populations as evolutionarily significant units with diverse genetic backgrounds could be an effective solution for present population recovery, with the possible relocation of individuals among different groups to help ensure future goral species prosperity.
Asunto(s)
Distribución Animal , Conservación de los Recursos Naturales , Genotipo , Rumiantes/genética , Animales , Femenino , Variación Genética , Masculino , Filogenia , Análisis para Determinación del Sexo/métodos , Análisis para Determinación del Sexo/veterinariaRESUMEN
The diversity of IoT services and applications brings enormous challenges to improving the performance of multiple computer tasks' scheduling in cross-layer cloud computing systems. Unfortunately, the commonly-employed frameworks fail to adapt to the new patterns on the cross-layer cloud. To solve this issue, we design a new computer task scheduling framework for multiple IoT services in cross-layer cloud computing systems. Specifically, we first analyze the features of the cross-layer cloud and computer tasks. Then, we design the scheduling framework based on the analysis and present detailed models to illustrate the procedures of using the framework. With the proposed framework, the IoT services deployed in cross-layer cloud computing systems can dynamically select suitable algorithms and use resources more effectively to finish computer tasks with different objectives. Finally, the algorithms are given based on the framework, and extensive experiments are also given to validate its effectiveness, as well as its superiority.
RESUMEN
Repetitive DNA, especially that due to transposable elements (TEs), makes up a large fraction of many genomes. Dfam is an open access database of families of repetitive DNA elements, in which each family is represented by a multiple sequence alignment and a profile hidden Markov model (HMM). The initial release of Dfam, featured in the 2013 NAR Database Issue, contained 1143 families of repetitive elements found in humans, and was used to produce more than 100 Mb of additional annotation of TE-derived regions in the human genome, with improved speed. Here, we describe recent advances, most notably expansion to 4150 total families including a comprehensive set of known repeat families from four new organisms (mouse, zebrafish, fly and nematode). We describe improvements to coverage, and to our methods for identifying and reducing false annotation. We also describe updates to the website interface. The Dfam website has moved to http://dfam.org. Seed alignments, profile HMMs, hit lists and other underlying data are available for download.