Bioinformatics analysis of transcriptional regulation of circadian genes in rat liver
© Nguyen et al.; licensee BioMed Central Ltd. 2014
Received: 8 November 2012
Accepted: 6 March 2014
Published: 25 March 2014
The circadian clock is a critical regulator of biological functions controlling behavioral, physiological and biochemical processes. Because the liver is the primary regulator of metabolites within the mammalian body and the disruption of circadian rhythms in liver is associated with severe illness, circadian regulators would play a strong role in maintaining liver function. However, the regulatory structure that governs circadian dynamics within the liver at a transcriptional level remains unknown. To explore this aspect, we analyzed hepatic transcriptional dynamics in Sprague-Dawley rats over a period of 24 hours to assess the genome-wide responses.
Using an unsupervised consensus clustering method, we identified four major gene expression clusters, corresponding to central carbon and nitrogen metabolism, membrane integrity, immune function, and DNA repair, all of which have dynamics which suggest regulation in a circadian manner. With the assumption that transcription factors (TFs) that are differentially expressed and contain CLOCK:BMAL1 binding sites on their proximal promoters are likely to be clock-controlled TFs, we were able to use promoter analysis to putatively identify additional clock-controlled TFs besides PARF and RORA families. These TFs are both functionally and temporally related to the clusters they regulate. Furthermore, we also identified significant sets of clock TFs that are potentially transcriptional regulators of gene clusters.
All together, we were able to propose a regulatory structure for circadian regulation which represents alternative paths for circadian control of different functions within the liver. Our prediction has been affirmed by functional and temporal analyses which are able to extend for similar studies.
The circadian clock is one of the most critical biological regulators for all living entities controlling behavioral, physiological and biochemical processes . Acting like a multifunctional timer with a roughly 24-h cycle , the circadian clock offers fitness advantages by endowing organisms with anticipatory mechanisms predicting periodic events, as opposed to responding in a reactive way to external signals . Robust circadian regulation is associated with fitness, while abnormal rhythms which are characteristic of stress are often linked with illness [4, 5]. As an example, metastatic colorectal cancer patients with severe alterations of the rest-activity circadian rhythm were predicted to have a fivefold increase in the risk of death compared to a normal rest-activity pattern .
In mammals, the circadian clock is composed of auto-regulated transcription-translation feedback loops whose primary elements are transcription factors (TFs) including the basic helix-loop-helix transcription factor CLOCK and the aryl hydrocarbon receptor nuclear translocator-like BMAL1, which drive the expression of period (PER) and cryptochrome (CRY). PER and CRY form the negative limb of the feedback loop, repressing the expression of CLOCK–and BMAL1–induced transcription . Evidence suggests that circadian gene expression is achieved through a transcriptional cascade in which the expression of other TFs responsible for the majority of rhythmic tissue-specific expression are primarily driven directly by the main clock transcription factor complex CLOCK:BMAL1 . Recently, this hypothesis has been reinforced by experiments showing that TFs whose expression patterns exhibit daily oscillation are direct targets of the CLOCK:BMAL1 complex [9, 10]. However, those studies only explored the transcriptome with one main clock transcription factor mutated and thus the entire picture of how the system behaves or how target genes are controlled is still a question. Consequently, among the aims of our study is to gain insights into the broader transcriptional regulation of circadian rhythms.
Liver is one of the most important organs regulating the metabolic activity of the body as well as producing acute phase proteins (APP, class of proteins responding to inflammatory stressors) in response to external stress . Genome-wide transcriptome-mapping studies have revealed that the fraction of diurnally regulated liver transcripts could amount up to 10%, with genes exhibiting circadian rhythms participate in various functions e.g. cell signaling, energy metabolism, amino acid, lipid and cholesterol metabolism, carbohydrate transport and metabolism, DNA replication, protein synthesis, signal transduction mechanisms, and immune response [8, 12–15]. Due to the significance of the circadian periodicity and the severe outcome of disorders following its disruption, a global perspective of the transcriptional dynamics and regulatory structure of circadian genes in liver may offer a molecular framework for studies on the hepatic circadian rhythm gene regulation.
In this study, we address two critical questions including (1) whether we can identify and functionally characterize circadian-regulated genes in an unsupervised manner, and (2) if the potential transcriptional regulatory mechanisms of circadian rhythms can be hypothesized. In order to obtain the necessary samples for analysis, animals were sacrificed at 9 am, 11 am, 1 pm, 5 pm, 1 am, and 9 am (~0, 2, 4, 8, 16, and 24 hr zeitgeber time) via exsanguinations and livers were harvested, allowing for the acquisition of transcriptional data from microarray readouts for those time points. Utilizing our previous consensus clustering  and techniques in promoter analysis , our results actually demonstrate that clusters of co-expressed genes do vary in a circadian manner and contain functionalities consistent with previous studies including immune function, cell repair, metabolism, and DNA repair. Additional clock-controlled TFs are also predicted. Interestingly, we observed that many clock-controlled TFs relevant to the regulation of genes within a circadian pattern show similar transcriptional dynamics. Furthermore, the predicted regulatory network and regulatory motifs of circadian patterns are shown to be consistent with the biological significance of the corresponding gene set and the circadian time of the transcriptional pattern. Finally, the identified transcription factors are intimately related to innate immunity and response to infection, providing further evidence that circadian rhythms play an important role in recovery from injury. While it is understood that recent evidence points to the important role alternative processes play for driving the induction of oscillations [18–20], the proposed approach presents a generalizable framework for assessing the regulatory structure underlying the transcriptional sources of oscillations.
Circadian dynamics and functions
The detailed dynamics of the four transcription patterns are shown in Figure 1 with the average expression profiles of all probesets in each corresponding cluster. Functional annotation was performed to develop a succinct description of the key roles of each pattern. In brief, pattern 1 (132 genes) is composed of genes whose activity peaks in the morning (ZT0 ~ 9:00 am) and is enriched by functions associated with metabolism, including (i) amino acid metabolism such as nitrogen metabolism (car14, car5a, cth, gls2), alanine, aspartate and glutamate metabolism (abat, agxt2, gls2, loc641316), glycine, serine and threonine metabolism (agxt2, cth, gldc), arginine and proline metabolism (agmat,gls2), beta-alanine metabolism (abat, dpys), and (ii) carbohydrate metabolism such as glycosaminoglycan degradation (gns, gusb), propanoate metabolism (abat,acacb), starch and sucrose metabolism (gusb, pygl). In addition, many other genes related to primary metabolism and are known to oscillate in a circadian manner e.g. bmal, gcjr, dhtkd1, cth, gaba, gldc, kynu, akr1c18, chka and agxt.
Pattern 2 contains 42 genes whose expression levels reach the maximum during the day time (ZT2 ~ 11:00 am) and appears to approach baseline before the dark phase. This pattern consists of genes with diverse functions including energy production (ppargc1a, coq10b), cell-cell junction (tanc1, epb4.1 l5), and membrane structure (ubr4, tmem106b). The genes associated with cell junction and component of membrane play a critical role in the integrity of the endothelial barrier function. Additionally, this pattern also contains several transcription factors involved in regulating circadian dynamics e.g. CREM (a transcription factor involved in output clock function and melatonin synthesis in the mouse ), KLF9 . In a similar manner, 46 genes in pattern 3 display a maximum activity during the light phase and are primarily composed of genes associated with the immune system. Genes in this temporal class are involved in the Chemokine signaling pathway (chuk, gnai3, stat3), RIG-I-like receptor signaling pathway (chuk, ddx3x), and Toll-like receptor signaling pathway (chuk, myd88). Also, there are numerous genes that were shown to have circadian dynamics and related to immune function in previous studies such as btg1, sfpq, and fkbp5. Finally, pattern 4 (70 genes) exhibits a reduced activity during the day time and reaches its peak activity around midnight. It is characterized by functions associated with DNA replication and repair, such as mismatch repair (rfc2, rfc5), DNA replication (rfc2, rfc5), and non-homologous end joining (rad50). Furthermore, timeless and calm1 are also known as genes with circadian rhythmicity that vary within this pattern.
Putative transcriptional regulators controlling circadian dynamics
Putative transcriptional regulators controlling circadian-relevant genes
Putative transcriptional regulators
63 genes / 132 genes*
AP1R, AP4R, CAAT, CLOX, CREB, E2FF, ETSF, EVI1, FKHD, GATA, GREF, HOXF, MYBL, MYT1, NKXH, NR2F, OCT1, PLAG, RUSH, RXRF, SORY, STAT, XBBF
15 genes / 42 genes
AHRR, CLOX, CREB, CTCF, E2FF, EREF, ETSF, EVI1, FKHD, GCMF, GLIF, HAND, HEAT, HOMF, HOXF, MYBL, NKXH, NR2F, OCT1, PARF, PAX5, PERO, RXRF, SORY, SP1F, XBBF
24 genes / 46 genes
AP1R, CREB, E2FF, EGRF, ETSF, GATA, HEAT, HOXF, IRFF, MYBL, NKXH, NR2F, PERO, RXRF, STAT, TBPF
32 genes / 70 genes
CREB, E2FF, ETSF, EVI1, NR2F, RXRF, SORY, SP1F
Transcriptional patterns of clock-controlled TFs and their temporal organization
Putative clock TFs regulating the transcription of circadian-relevant genes
Putative clock TFs involved in the transcriptional regulation of circadian-relevant genes
Sets of regulators
p-value | q-value
Sets of regulators
p-value | q-value
1 109/132 genes*
CREB, RORA, STAT
0.0078 | 0.0616
FKHD, PARF, RXRF
0.0377 | 0.0877
CREB, RORA, RXRF
0.0116 | 0.0616
RORA, RXRF, STAT
0.0409 | 0.0877
0.0191 | 0.0753
CREB, FKHD, RORA
0.0473 | 0.0877
2 27/42 genes
AHRR, CREB, STAT
0.0077 | 0.0616
0.0303 | 0.0877
AHRR, CREB, EBOX
0.0079 | 0.0616
AHRR, RXRF, STAT
0.0327 | 0.0877
AHRR, CREB, KLFS, RXRF
0.0130 | 0.0616
AHRR, KLFS, STAT
0.0482 | 0.0877
3 39/46 genes
CREB, EBOX, RXRF, STAT
0.0428 | 0.0877
4 59/70 genes
Identified circadian dynamics and biological significance
The liver expresses a diverse set of genes and previous research indicated that many hepatic genes are under direct or indirect circadian control . Despite the fact that unsupervised pattern identification methods were used to identify dominant co-expressed gene sets, our results show that selected genes oscillate in a circadian manner, as shown in Figure 1. Functional characterization reveals that enriched functions are associated with metabolism, energy production, immune system and DNA replication and repair.
The liver plays an important role in the orchestration of metabolic functions such as the regulation of metabolic fuels, protein synthesis, iron and vitamin storage, as well as nutrient intake . As might be expected, animals subjected to a light–dark schedule display pronounced rhythms in glycogen content, with a peak occurring late in the night, following the main period of food intake. This is also in agreement with genes identified in pattern 1, where enriched metabolic functions, especially glucose metabolism, reach their peak during the dark phase. Since the dark phase represents the active period for rats, where they are moving and eating, this increase in metabolism is indicative of both energy burn to fuel muscle usage, as well as digestive effects on the liver, and an influx of glucose and other amino acids to be processed through glycogen production and the urea cycle. Because the current study does not differentiate from fed animals and fasted animals, it is impossible to determine how much increased metabolic activity is due to demand (movement), versus supply (digestion). In addition, it has been suggested that the expression patterns of circadian rhythms in peripheral organs such as the liver are the result of delayed feedback loops of gene products from circadian regulators, suggesting that upstream regulation in addition to direct transcription factor binding can occur up to 6 hours prior to the peak of gene expression .
Energy production is also time-of-day dependent. It is reported that the production of ATP in rat liver is also regulated by the circadian rhythm and is lower in the dark phase . Membrane integrity and cell to cell junctions are another critical circadian function that has been previously hypothesized to be important during rest , and contain significant circadian regulation . Circadian pattern 2 is made up of a diverse array of genes whose functions are related to energy production and membrane structure, but are up regulated during the day. Many of these metabolic genes are associated with mitochondrial biogenesis specifically, which is a process that has been previously shown to occur during the resting phase of the circadian cycle . As rats are nocturnal, their day cycle is indicative of their sleep cycle, this is in prior agreement and it suggests that mitochondrial biogenesis is a major liver function during the rest portion of their circadian rhythm. Genes that are associated with apoptosis are also found within this pattern such asbclaf1, rarb, and dapk1 and are further indication of cellular repair–cells which are too damaged to be simply repaired are killed and replaced in order to prepare the host for another day of activity.
Immune system functions, including lymphocyte proliferation, natural killer cell activity, humoral immune response, absolute and relative numbers of circulating white blood cells and their subsets, cytokine levels, and serum cortisol, are known to be entrained by circadian cues . Also, it has been previously reported that TNFα secretion was significantly increased in burn vs. sham isolated splenocytes cells only when injury took place in the morning  further verifying that the response to injury is time-of-day dependent as a result of the circadian variability of the immune functions. This is confirmed by the selected gene set exhibiting pattern 3 includingbtg1, sfpq, fkbp5, which is primarily related to immune function, and has consistent dynamics with previous works [22, 26, 27]. However, further analysis of the dynamics observes a peak in immune function at the end of the day (ZT8 ~ 5:00 pm), which is the time rats are beginning to awaken. Over the course of the night, when the rats are active, immune function gradually decreases, until it reaches a minimum when the night ends, and the rats rest. This implies that there may be a recovery mechanism associated with sleep that allows for the rejuvenation of immune function in order to ward off infection during active periods.
The strategy of adaptive circadian clocks could be also timing of UV-sensitive cellular processes to occur at night to avoid UV-induced damage . Our analysis showed that functions related to DNA replication/repair (pattern 4) achieve a maximum activity around midnight, which is consistent with previous works . However, unlike the previous patterns, the circadian dynamics of these genes are not dependant on the habits of the host animal, but rather upon external conditions related to the environment. Unlike the other patterns, the genes contained in pattern 4 do not map to specific transcription factors directly, which suggests that this group is under a different form of control than the others. The expression pattern of pattern 4 is similar to the late cry1 phase of the circadian rhythm , which were identified to be under the control of newly identified D boxes that overlap with E-boxes to provide delayed expression of genes which contributes to rhythmicity. Furthermore, the cry1 expression pattern, whose phase resembles cluster 4, is heavily dependent on other circadian regulators for its robustness , which suggests that pattern 4 may be controlled through upstream transcription factor regulators through their gene products, in addition to direct TF binding.
Putative regulatory motifs of circadian-relevant genes
Simplified networks are shown in Figure 3b, which simply condense the transcription factors into three categories: core clock transcription factors (CLOCK:BMAL1), clock controlled transcription factors (ccTFs) and other transcription factors. Genes which exhibit transcription pattern 2 show the most circadian influence, with the corresponding promoter set containing enriched binding sites for the core clock TFs, ccTFs, and other TFs. Meanwhile genes following pattern 4 has the least circadian influence, being only controlled indirectly through two levels of regulation by the core clock transcription factors, likely through their gene products . Because of the overlapping levels of regulation associated with pattern 2, it can be expected that disruptions to the circadian cycle will be less effective at disrupting its gene expression, as those layers of regulation could potentially provide redundancy that preserves circadian gene expression. Meanwhile, pattern 4, can be putatively said to be more sensitive to circadian disruptions, due to the lack of backup regulation. This is supported by the fact that pattern 4 has a similar expression to the cry1 gene, which is a core circadian regulator known for its sensitivity to changes in the overall circadian rhythm. Further, since pattern 2 is representative of host repair during rest, it makes sense that it would be protected against disruption: proper sleep is critical to health and disease recovery , and the redundancy of the host repair genes may reflect that need. Similarly, pattern 4 is associated with DNA repair, which has been shown to be severely affected following injury due to the increased presence of reactive oxygen species [44, 45], known to damage DNA. Thus the putative sensitivity of this pattern to circadian disruption may be a result of altered DNA repair patterns during periods of stress. Genes in pattern 1 and 3 are both under similar circadian regulation, despite being significantly separated in circadian time, as discussed previously. These patterns primarily relate to metabolism and innate immunity, two functions which have been shown to be closely linked through a phenomenon known as hypermetabolism . These functions show significant circadian sensitivity, and because of their peaks at different circadian times (ZT0 and ZT8), it is possible that due to different innate immune and metabolic states over the course of the circadian day, the time of injury can significantly impact the severity of the immune response and the onset of hypermetabolism. The hypothesis that circadian rhythms strongly impact immune responses following injury has been postulated before , but further experiments are needed in order to conclusively link the functions, and dynamics of genes in pattern 1 & 3 to injury responsive elements.
Dynamic and functional agreement between clock-controlled TFs and their associated gene sets
In Figure 2, it is worth noting that the transcription factors that have early circadian time peaks (ZT0 to ZT4), namely STAT5A, FOXO3, FOXN3, CREM, and RARB are all associated with the regulation of genes in transcription pattern 1, whose genes also peak very early in circadian time. Transcription factors which peak at somewhat later times (ZT4-ZT8) include RARB, KLF9, AHR, ATF6, and STAT3, which have all been associated with pattern 2. This pattern peaks at a later time period than the first pattern, and this is reflected in the shift in circadian time peaks in its transcriptional regulators. Finally, the transcription factors which peak at the latest circadian times (ZT8-Z16) include ATF6, STAT3, and NR1I3, which are all associated with pattern 3. This pattern peaks at the latest time point following the diurnal onset, and again, this is reflected in the shifted circadian time peaks of its regulators. In general, many of transcription factors control the same set of genes exhibiting a circadian pattern express a very similar circadian profile.
Though many transcription factors have been identified, the three clock-controlled TF families which control the functions of all of the patterns through binding sites on non-clock controlled TFs are FKHD, KLFS, and STAT (Figure 3). The FKHD transcription factor family includes the Fox-O transcription factor, which is well known for its protective role against oxidative stress within the body . Similarly, KLFS is well known for its role in the regulation of anti polymicrobial bacteria activity within the body and metabolism in liver [25, 49], while STAT has already been identified as a critical transcription factor during the outcome of sepsis . In particular, KLF has been strongly associated with liver metabolism in particular , which is reinforced by its regulation of cluster 2, which has many metabolic genes associated with it. Because KLF is strongly associated with CLOCK and BMAL, its presence as a regulator of these oscillatory gene clusters suggests significant circadian regulation of these functions. Together these three mediators and their functions indicate that the overarching control of clock TFs to the patterns indirectly through non clock-controlled TFs is controlled primarily by the innate immune response, and may serve to coordinate the metabolic functions and toll like receptor functions found in the patterns in the event of an innate immune response.
A TF family that is considered as a common regulator of genes across all four circadian patterns is CREB. It has been previously recognized to be prototypical stimulus induced in function , and therefore plays a wide variety of roles in different tissues. However, it has been strongly linked to functional circadian regulation throughout the body, and thus it is not surprising that CREB regulatory elements are found in all four patterns, both directly and indirectly regulated by CLOCK:BMAL1. Unique clock controlled TF families that regulate genes in pattern 1 are PARF and RORA. PARF includes well known clock TFs (e.g. DPB, TEF) which are directly targets of CLOCK:BMAL1 [52, 53]. RORA is another well known circadian transcription factor family, which has also been shown to play a role in metabolism in many tissues, including the muscle and liver systems [54, 55]. Additionally, RXRF transcription factors represent a family of nuclear receptors that target a variety of signaling pathways, which are primarily related to metabolism, cell differentiation, and cell death . Thus, much like others associated with pattern 1, RXRF factor family has significantly overlapping functions in the literature to the genes that they control. Unique transcription factor families that regulate genes in patterns 2 and 3 include AHRR and EBOX. AHRR, which is a unique clock transcription factor family associated with pattern 2, has been associated with the regulation of cell adhesion and matrix formation , which is in agreement with the functions of genes in pattern 2. EBOX, which is associated with the regulation of genes in both patterns, has implicit functions in innate immunity, and it has been shown that the mechanism by which TNF-α, a well known mediator of the inflammatory response, suppresses clock genes is through interference with EBOX mediated transcription . EBOX transcription factors have also been shown to be key regulators of the circadian rhythm, including well known circadian genes and clock controlled transcription factors . Thus, both EBOX and AHRR not only have functions that are similar to those of their respective gene sets, but also have prior association with circadian regulation, which suggests that many of the regulators which control innate immunity and metabolism within the liver are at least partially under the regulation of the circadian rhythm.
Alternative factors relevant to the regulation of circadian genes
Our promoter analysis of transcription factors is based on the assumption that clock TFs are mainly responsible for the transcriptional dynamics that are observed over 24 hours within the liver. However, other factors, including SCN stimuli and food intake in the form of releasing glucocorticoids from the adrenal gland, light stimuli, and heat, have been known to impact circadian rhythms . This is particularly important in the context of genes in pattern 4, whose functions are closely related to DNA repair and maintenance and which peaks during the night, thereby repairing any UV damage that occurs.
Among transcription factor families identified from Table 1, GREF represents a known transcription factor responsive element that is stimulated by glucocorticoids . Glucocorticoids are well known molecules that are coordinated by the SCN in order to entrain peripheral tissues in a circadian manner . Since genes in pattern 1 are controlled by this transcription factor family, it suggests that many liver metabolic functions may also be regulated by the SCN. This fits the current paradigm of SCN regulation, since the SCN has been associated with the regulation of feeding patterns . Therefore, by controlling metabolic enzymes within the liver to be most active at ZT0, the SCN can coordinate metabolic functions to coincide with feeding patterns.
Another alternative regulatory mechanism for circadian rhythms is through light. Although there is no direct light input to liver, light may trigger a signaling cascade which eventually activates some functional TFs in liver. This activation requires the aid of protein kinases (PKA) , which activate CREB and allow it to translocate to the nucleus for transcriptional regulation . This two-step activation controls all four circadian pattern motifs through CREB factors, and may, in addition to CLOCK:BMAL1, be responsible for the diurnal activity observed within the liver. The importance of light in the sleep/wake cycle has been well established , and thus it is not surprising that the circadian activity of the liver may be entrained to the same light/dark rhythms that govern sleeping cycles in mammals.
Microarrays were used to characterize the gene expression within rat livers over 24 hours, and various computational tools have been applied to identify and distribute differentially expressed genes from the microarray data into four distinct patterns. Pattern 1 is primarily related to central metabolism with the peak at circadian time ZT0. Pattern 2 is related to cell repair and cell junction formations, and peaks at circadian time ZT4. Pattern 3 is primarily related to the innate immune response and peaks at circadian time ZT8, while Pattern 4 is related to DNA repair and peaks at circadian time ZT16. The study also proposed a unique regulatory structure for circadian rhythms in the liver, which aims to predict the ways in which circadian patterns can potentially impact injury. By using techniques from promoter analysis coupled with the proposed hypothesis of clock TF regulation, a set of transcription factors and putative regulatory structure have been established which show to be consistent in agreement with functional and temporal annotations. These transcription factor dynamics appear to play a strong role in circadian regulation, and therefore potentially impact the liver’s response to environmental challenges at different circadian times. In conclusion, our study aims at (1) unsupervised identification of circadian-relevant genes and (2) exploration of potential regulatory mechanisms relevant to the transcriptional regulation of circadian genes. While doing that, we eventually proposed a promising computational framework that can predict relevant transcriptional factors and putative regulatory structure of circadian genes in a genome-wide manner, establishing a platform towards large-scale comparisons of the regulation in homeostatic systems across multiple tissues and/or conditions.
Male Sprague–Dawley rats (Charles River Labs, Wilmington, MA) weighing between 150 and 200 g were utilized for this study. The animals were housed in a temperature-controlled environment (25°C) with a 12-hour light–dark cycle and provided water and standard chow ad libitum. Because the animals are nocturnal, this establishes a 12 hour LD cycle for the animals, where ZT0-ZT12 represents the resting portion of the day, while ZT12-ZT0 represents the active portion of the day. All experimental procedures were carried out in accordance with National Research Council guidelines and approved by the Rutgers University Animal Care and Facilities Committee.
Animals are sacrificed at 9 am, 11 am, 1 pm, 5 pm, 1 am, and 9 am (~0, 2, 4, 8, 16, and 24 hr zeitgeber time). Animals were originally used as part of the study by Yang et al.  to analyze gene expression following burn injury, which dictates the choice of time points. Ideally, future works would aim at sampling the latter parts of the night and day phases equally to observe trends throughout the time course. Liver tissues were collected and frozen for microarray analysis (n = 3 per time point per group). The tissues were lysed and homogenized using Trizol, and the RNAs were further purified and treated with DNase using RNeasy columns (Qiagen). Then cRNAs prepared from the RNAs of liver tissues using protocols provided by Affymetrix were utilized to hybridize Rat Genome 230 2.0 Array (GeneChip, Affymetrix) comprised of 31,099 probe sets.
Microarray data analysis
Genome expression data analysis included normalization, filtering for differential expression, and clustering. DNA chip analyzer (dChip) software  was used with invariant-set normalization and perfect match (PM) model to generate expression values. The analysis of variance (ANOVA) test implemented in R  was applied to filter differentially expressed probesets in the dataset with p-value < 0.001. At this step, to control the false discovery rate the method proposed by Storey  was used with a positive false discovery rate q-value cutoff of 0.01. Due to a limitation of data, such strict thresholds were used and thus several well-known circadian genes were not present in the filtered gene set (see Additional file 1, sheet ‘filtering’). Subsequently, consensus clustering  was applied in order to identify, in an unsupervised manner, coherent clusters of co-expressed genes with p-value < 0.001. The biological relevance of the intrinsic responses was also characterized by evaluating the enrichment of the corresponding subsets in circadian rhythm specific pathways using KEGG database through ARRAYTRACK (p-value < 0.05)  and characterizing functionally individual genes.
Promoter extraction and processing
Promoters of genes including all transcript-relevant alternative promoters were extracted from a rich database of promoter information. If there is no experimentally defined length suggested by Genomatix, a default length, 500 bp upstream and 100 bp downstream of the transcription start site which is the most enriched region of transcription factor binding sites (TFBSs) according to Genomatix, is used . In order to accelerate the process of identifying putative transcriptional regulators, promoters are pre-processed as in . Specifically, MatInspector  is applied to scan for PWM matches on those promoter sequences using optimal parameters from MatBase which ensures that the minimum number of matches found in non-regulatory sequences i.e. the false positive matches is minimized . Each promoter of a gene is then re-modelled to become a list of TFBSs ordered by their local positions on the promoter sequences and represented by corresponding TF names along with their binding orientations. The conversion supports for fast search the presence of a TFBS or a set of TFBSs on promoter sequences .
Prediction of relevant transcriptional regulators
In order to predict putative TFs relevant to the transcriptional regulation of circadian-relevant gene sets, we explored the basic underlying assumption of comparative genomics which states that functional regions evolve in a constrained fashion and therefore at a lower rate than non-functional regions [41, 42]. Consequently, orthologous promoters of each gene are further extracted. Each promoter P is now characterized by a set of orthologous promoters from the same gene of other vertebrate species, if available (e.g. Homo sapiens, Macacamulatta, Pantroglodytes, Musmusculus, Equuscaballus, Canis lupus familiaris, and Bos Taurus). To be consistent in the search for conserved regions on promoter sequences in order to identify putative TFBSs, we eliminate those that do not consist of more than two orthologous promoters . DiAlign TF  with default parameters (core similarity: 0.75, matrix similarity: optimal threshold for each PWM suggested from MatBase) was applied to identify conserved regions on promoter P. We next apply MatInspector  to scan for all physical TFBSs and only keep those that locate on conserved regions. Subsequently, given a set of genes, TFBSs that are enriched with a common threshold (70% in this study) are identified and considered as TFs relevant to the transcriptional regulation of the gene set.
Identification of clock-controlled transcription factors
Besides TF families composing of clock TFs identified from literature (CLOCK:BMAL1, PARF and RORA ), we hypothesized that transcription factors that contain CLOCK:BMAL1 binding sites on their promoters and also differentially expressed in the homeostatic system are likely to be clock-controlled TFs. After identification of differentially expressed genes from the entire dataset, MatInspector  was applied to scan for TFBSs present on corresponding promoters of each genes. Genes that contain CLOCK:BMAL1 binding sites and are also reported as transcription factors in MatBase  are considered as clock-controlled TFs.
Identification of common sets of clock transcription factors
It has been noticed that TFs in higher eukaryotes regulate gene expression in a combinatorial manner rather than in isolation [31, 32]. Consequently, to explore the underlying mechanisms of clock TFs in the relation to the regulation of these circadian rhythms, we hypothesize that significant sets of clock TFs present on promoter sets of co-expressed circadian-relevant genes are more likely to be coordinated regulators in modulating the transcriptional process of circadian rhythms. We thus applied the breadth first search technique to identify set of clock TFs that are commonly present on the promoter set of genes in each circadian pattern. Since a gene may have multiple alternative promoters, if a set of clock TFs present on any promoter of the gene it is considered as presence on that gene. The procedure first identifies all potential clock TFBSs that are commonly present on the corresponding promoter set and then search for all possible combinations of those TFBSs (common level of 70%). Statistical significance of each common set of clock TFs is evaluated to select significant sets for further examination (p-value < 0.05, q-value < 0.10) (detailed in Additional file 1).
Statistical significance of sets of clock TFs
In order to construct the relations of clock TFs to the regulation of circadian rhythms, we estimate the significance of common sets of clock TFs identified from the corresponding gene set of each transcription pattern vs. the background set. The procedure proposed by Bozek et al.  was followed to extract a GC-matched background gene set (~10,000 genes) in order to compensate the relatively high GC-content promoters of circadian relevant genes (detailed in Additional file 1).
where B and b is the number of genes and the number of hits respectively in the background set; N and n is the number of genes and hits in gene set G, respectively. A ‘hit’ is the presence of set A on a promoter of a gene.
Additionally, to restrain the multiple testing problem in searching for significant sets of clock TFs, the bootstrap method proposed by Storey  was utilized to estimate the positive false discovery rates (q-value). A rate of 10% was used in this case.
The authors gratefully acknowledge the financial support from NIH grant GM082974.
National Institute of Health (NIH) GM082974.
- Rana S, Mahmood S: Circadian rhythm and its role in malignancy. J Circadian Rhythms. 2010, 8: 3-10.1186/1740-3391-8-3.View ArticlePubMed CentralPubMedGoogle Scholar
- Koyanagi S, Ohdo S: Alteration of intrinsic biological rhythms during interferon treatment and its possible mechanism. Mol Pharmacol. 2002, 62 (6): 1393-1399. 10.1124/mol.62.6.1393.View ArticlePubMedGoogle Scholar
- Vitalini MW, de Paula RM, Park WD, Bell-Pedersen D: The rhythms of life: circadian output pathways in Neurospora. J Biol Rhythms. 2006, 21 (6): 432-444. 10.1177/0748730406294396.View ArticlePubMedGoogle Scholar
- Lowry SF: The stressed host response to infection: the disruptive signals and rhythms of systemic inflammation. Surg Clin North Am. 2009, 89 (2): 311-326. 10.1016/j.suc.2008.09.004. viiView ArticlePubMed CentralPubMedGoogle Scholar
- Filipski E, Li XM, Levi F: Disruption of circadian coordination and malignant growth. Cancer Causes Control. 2006, 17 (4): 509-514. 10.1007/s10552-005-9007-4.View ArticlePubMedGoogle Scholar
- Mormont MC, Waterhouse J, Bleuzen P, Giacchetti S, Jami A, Bogdan A, Lellouch J, Misset JL, Touitou Y, Levi F: Marked 24-h rest/activity rhythms are associated with better quality of life, better response, and longer survival in patients with metastatic colorectal cancer and good performance status. Clin Cancer Res. 2000, 6 (8): 3038-3045.PubMedGoogle Scholar
- Lowrey PL, Takahashi JS: Mammalian circadian biology: elucidating genome-wide levels of temporal organization. Annu Rev Genomics Hum Genet. 2004, 5: 407-441. 10.1146/annurev.genom.5.061903.175925.View ArticlePubMed CentralPubMedGoogle Scholar
- Duffield GE, Best JD, Meurers BH, Bittner A, Loros JJ, Dunlap JC: Circadian programs of transcriptional activation, signaling, and protein turnover revealed by microarray analysis of mammalian cells. Curr Biol. 2002, 12 (7): 551-557. 10.1016/S0960-9822(02)00765-0.View ArticlePubMedGoogle Scholar
- Oishi K, Miyazaki K, Kadota K, Kikuno R, Nagase T, Atsumi G, Ohkura N, Azama T, Mesaki M, Yukimasa S, Kobayashi H, Iitaka C, Umehara T, Horikoshi M, Kudo T, Shimizu Y, Yano M, Monden M, Machida K, Matsuda J, Horie S, Todo T, Ishida N: Genome-wide expression analysis of mouse liver reveals CLOCK-regulated circadian output genes. J Biol Chem. 2003, 278 (42): 41519-41527. 10.1074/jbc.M304564200.View ArticlePubMedGoogle Scholar
- Grechez-Cassiau A, Panda S, Lacoche S, Teboul M, Azmi S, Laudet V, Hogenesch JB, Taneja R, Delaunay F: The transcriptional repressor STRA13 regulates a subset of peripheral circadian outputs. J Biol Chem. 2004, 279 (2): 1141-1150.View ArticlePubMedGoogle Scholar
- Yang E, Maguire T, Yarmush ML, Berthiaume F, Androulakis IP: Bioinformatics analysis of the early inflammatory response in a rat thermal injury model. BMC Bioinforma. 2007, 8: 10-10.1186/1471-2105-8-10.View ArticleGoogle Scholar
- Akhtar RA, Reddy AB, Maywood ES, Clayton JD, King VM, Smith AG, Gant TW, Hastings MH, Kyriacou CP: Circadian cycling of the mouse liver transcriptome, as revealed by cDNA microarray, is driven by the suprachiasmatic nucleus. Curr Biol. 2002, 12 (7): 540-550. 10.1016/S0960-9822(02)00759-5.View ArticlePubMedGoogle Scholar
- Panda S, Antoch MP, Miller BH, Su AI, Schook AB, Straume M, Schultz PG, Kay SA, Takahashi JS, Hogenesch JB: Coordinated transcription of key pathways in the mouse by the circadian clock. Cell. 2002, 109 (3): 307-320. 10.1016/S0092-8674(02)00722-5.View ArticlePubMedGoogle Scholar
- Storch KF, Lipan O, Leykin I, Viswanathan N, Davis FC, Wong WH, Weitz CJ: Extensive and divergent circadian gene expression in liver and heart. Nature. 2002, 417 (6884): 78-83. 10.1038/nature744.View ArticlePubMedGoogle Scholar
- Walker JR, Hogenesch JB: RNA profiling in circadian biology. Methods Enzymol. 2005, 393: 366-376.View ArticlePubMedGoogle Scholar
- Nguyen TT, Nowakowski RS, Androulakis IP: Unsupervised selection of highly coexpressed and noncoexpressed genes using a consensus clustering approach. OMICS. 2009, 13 (3): 219-237. 10.1089/omi.2008.0074.View ArticlePubMedGoogle Scholar
- Nguyen TT, Androulakis PI: Recent advances in the computational discovery of transcription factor binding sites. Algorithms. 2009, 2 (1): 582-605. 10.3390/a2010582.View ArticleGoogle Scholar
- Naef F: Circadian clocks go in vitro: purely post-translational oscillators in cyanobacteria. Mol Syst Biol. 2005, 1: 0019-2005View ArticlePubMedGoogle Scholar
- Gao P, Yoo SH, Lee KJ, Rosensweig C, Takahashi JS, Chen BP, Green CB: Phosphorylation of the Cryptochrome 1C-terminal tail regulates circadian period length. J Biol Chem. 2013, 288 (49): 35277-35286. 10.1074/jbc.M113.509604.View ArticlePubMed CentralPubMedGoogle Scholar
- Li Y, Rosbash M: Accelerated degradation of perS protein provides insight into light-mediated phase shifting. J Biol Rhythms. 2013, 28 (3): 171-182. 10.1177/0748730413489797.View ArticlePubMedGoogle Scholar
- Wichert S, Fokianos K, Strimmer K: Identifying periodically expressed transcripts in microarray time series data. Bioinformatics. 2004, 20 (1): 5-20. 10.1093/bioinformatics/btg364.View ArticlePubMedGoogle Scholar
- Almon RR, Yang E, Lai W, Androulakis IP, DuBois DC, Jusko WJ: Circadian variations in rat liver gene expression: relationships to drug actions. J Pharmacol Exp Ther. 2008, 326 (3): 700-716. 10.1124/jpet.108.140186.View ArticlePubMed CentralPubMedGoogle Scholar
- Hirao J, Niino N, Arakawa S, Shibata S, Mori K, Ando Y, Furukawa T, Sanbuissho A, Manabe S, Mori Y, Nishihara M: Circadian modulation of hepatic transcriptome in transgenic rats expressing human growth hormone. J Toxicol Sci. 2010, 35 (5): 673-685. 10.2131/jts.35.673.View ArticlePubMedGoogle Scholar
- Cermakian N, Sassone-Corsi P: Multilevel regulation of the circadian clock. Nat Rev Mol Cell Biol. 2000, 1 (1): 59-67. 10.1038/35036078.View ArticlePubMedGoogle Scholar
- Guillaumond F, Grechez-Cassiau A, Subramaniam M, Brangolo S, Peteri-Brunback B, Staels B, Fievet C, Spelsberg TC, Delaunay F, Teboul M: Kruppel-like factor KLF10 is a link between the circadian clock and metabolism in liver. Mol Cell Biol. 2010, 30 (12): 3059-3070. 10.1128/MCB.01141-09.View ArticlePubMed CentralPubMedGoogle Scholar
- Zambon AC, McDearmon EL, Salomonis N, Vranizan KM, Johansen KL, Adey D, Takahashi JS, Schambelan M, Conklin BR: Time- and exercise-dependent gene regulation in human skeletal muscle. Genome Biol. 2003, 4 (10): R61-10.1186/gb-2003-4-10-r61.View ArticlePubMed CentralPubMedGoogle Scholar
- Keller M, Mazuch J, Abraham U, Eom GD, Herzog ED, Volk HD, Kramer A, Maier B: A circadian clock in macrophages controls inflammatory immune responses. Proc Natl Acad Sci U S A. 2009, 106 (50): 21407-21412. 10.1073/pnas.0906361106.View ArticlePubMed CentralPubMedGoogle Scholar
- King DP, Takahashi JS: Molecular genetics of circadian rhythms in mammals. Annu Rev Neurosci. 2000, 23: 713-742. 10.1146/annurev.neuro.23.1.713.View ArticlePubMedGoogle Scholar
- Rodriguez-Caso C, Medina MA, Sole RV: Topology, tinkering and evolution of the human transcription factor network. Febs J. 2005, 272 (24): 6423-6434. 10.1111/j.1742-4658.2005.05041.x.View ArticlePubMedGoogle Scholar
- Ghorbel MT, Coulson JM, Murphy D: Cross-talk between hypoxic and circadian pathways: cooperative roles for hypoxia-inducible factor 1alpha and CLOCK in transcriptional activation of the vasopressin gene. Mol Cell Neurosci. 2003, 22 (3): 396-404. 10.1016/S1044-7431(02)00019-2.View ArticlePubMedGoogle Scholar
- Fessele S, Maier H, Zischek C, Nelson PJ, Werner T: Regulatory context is a crucial part of gene function. Trends Genet. 2002, 18 (2): 60-63. 10.1016/S0168-9525(02)02591-X.View ArticlePubMedGoogle Scholar
- Harbison CT, Gordon DB, Lee TI, Rinaldi NJ, Macisaac KD, Danford TW, Hannett NM, Tagne JB, Reynolds DB, Yoo J, Jennings EG, Zeitlinger J, Pokholok DK, Kellis M, Rolfe PA, Takusagawa KT, Lander ES, Gifford DK, Fraenkel E, Young RA: Transcriptional regulatory code of a eukaryotic genome. Nature. 2004, 431 (7004): 99-104. 10.1038/nature02800.View ArticlePubMed CentralPubMedGoogle Scholar
- Davidson AJ, Castanon-Cervantes O, Stephan FK: Daily oscillations in liver function: diurnal vs circadian rhythmicity. Liver Int. 2004, 24 (3): 179-186. 10.1111/j.1478-3231.2004.00917.x.View ArticlePubMedGoogle Scholar
- Korencic A, Bordyugov G, Kosir R, Rozman D, Golicnik M, Herzel H: The interplay of cis-regulatory elements rules circadian rhythms in mouse liver. PLoS One. 2012, 7 (11): e46835-10.1371/journal.pone.0046835.View ArticlePubMed CentralPubMedGoogle Scholar
- Robinson JL, Foustock S, Chanez M, Bois-Joyeux B, Peret J: Circadian variation of liver metabolites and amino acids in rats adapted to a high protein, carbohydrate-free diet. J Nutr. 1981, 111 (10): 1711-1720.PubMedGoogle Scholar
- Guo JS, Chau JF, Cho CH, Koo MW: Partial sleep deprivation compromises gastric mucosal integrity in rats. Life Sci. 2005, 77 (2): 220-229. 10.1016/j.lfs.2004.12.027.View ArticlePubMedGoogle Scholar
- Bass J, Takahashi JS: Circadian integration of metabolism and energetics. Science. 2010, 330 (6009): 1349-1354. 10.1126/science.1195027.View ArticlePubMed CentralPubMedGoogle Scholar
- Holzheimer RG, Curley P, Saporoschetz IB, Doherty JM, Mannick JA, Rodrick ML: Circadian rhythm of cytokine secretion following thermal injury in mice: implications for burn and trauma research. Shock. 2002, 17 (6): 527-529. 10.1097/00024382-200206000-00015.View ArticlePubMedGoogle Scholar
- Nikaido SS, Johnson CH: Daily and circadian variation in survival from ultraviolet radiation in Chlamydomonas reinhardtii. Photochem Photobiol. 2000, 71 (6): 758-765. 10.1562/0031-8655(2000)071<0758:DACVIS>2.0.CO;2.View ArticlePubMedGoogle Scholar
- Ukai-Tadenuma M, Yamada RG, Xu H, Ripperger JA, Liu AC, Ueda HR: Delay in feedback repression by cryptochrome 1 is required for circadian clock function. Cell. 2011, 144 (2): 268-281. 10.1016/j.cell.2010.12.019.View ArticlePubMedGoogle Scholar
- Doniger SW, Huh J, Fay JC: Identification of functional transcription factor binding sites using closely related Saccharomyces species. Genome Res. 2005, 15 (5): 701-709. 10.1101/gr.3578205.View ArticlePubMed CentralPubMedGoogle Scholar
- Hardison RC: Conserved noncoding sequences are reliable guides to regulatory elements. Trends Genet. 2000, 16 (9): 369-372. 10.1016/S0168-9525(00)02081-3.View ArticlePubMedGoogle Scholar
- Lapshina KV, Ekimova IV: Effects of sleep deprivation on measures of the febrile reaction and the recovery of somatovisceral functions and sleep in endotoxemia. Neurosci Behav Physiol. 2010, 40 (4): 381-388. 10.1007/s11055-010-9268-6.View ArticlePubMedGoogle Scholar
- Czaja MJ: Cell signaling in oxidative stress-induced liver injury. Semin Liver Dis. 2007, 27 (4): 378-389. 10.1055/s-2007-991514.View ArticlePubMedGoogle Scholar
- Jaeschke H: Reactive oxygen and mechanisms of inflammatory liver injury: Present concepts. J Gastroenterol Hepatol. 2011, 26 (Suppl 1): 173-179.View ArticlePubMedGoogle Scholar
- Cerra FB: Hypermetabolism, organ failure, and metabolic support. Surgery. 1987, 101 (1): 1-14.PubMedGoogle Scholar
- Arjona A, Sarkar D: Are circadian rhythms the code of hypothalamic-immune communication? insights from natural killer cells. Neurochem Res. 2008, 33 (4): 708-718. 10.1007/s11064-007-9501-z.View ArticlePubMedGoogle Scholar
- Choi J, Oh S, Lee D, Oh HJ, Park JY, Lee SB, Lim DS: Mst1-FoxO signaling protects Naive T lymphocytes from cellular oxidative stress in mice. PLoS One. 2009, 4 (11): e8011-10.1371/journal.pone.0008011.View ArticlePubMed CentralPubMedGoogle Scholar
- Mahabeleshwar GH, Kawanami D, Sharma N, Takami Y, Zhou G, Shi H, Nayak L, Jeyaraj D, Grealy R, White M: The myeloid transcription factor KLF2 regulates the host response to polymicrobial infection and endotoxic shock. Immunity. 2011, 34 (5): 715-728. 10.1016/j.immuni.2011.04.014.View ArticlePubMed CentralPubMedGoogle Scholar
- Andrejko KM, Chen J, Deutschman CS: Intrahepatic STAT-3 activation and acute phase gene expression predict outcome after CLP sepsis in the rat. Am J Physiol Gastrointest Liver Physiol. 1998, 275 (6): G1423-G1429.Google Scholar
- Lonze BE, Ginty DD: Function and Regulation of CREB Family Transcription Factors in the Nervous System. Neuron. 2002, 35 (4): 605-623. 10.1016/S0896-6273(02)00828-0.View ArticlePubMedGoogle Scholar
- Rey G, Cesbron F, Rougemont J, Reinke H, Brunner M, Naef F: Genome-wide and phase-specific DNA-binding rhythms of BMAL1 control circadian output functions in mouse liver. PLoS Biol. 2011, 9 (2): e1000595-10.1371/journal.pbio.1000595.View ArticlePubMed CentralPubMedGoogle Scholar
- Yan J, Wang H, Liu Y, Shao C: Analysis of gene regulatory networks in the mammalian circadian rhythm. PLoS Comput Biol. 2008, 4 (10): e1000193-10.1371/journal.pcbi.1000193.View ArticlePubMed CentralPubMedGoogle Scholar
- Yang X, Downes M, Yu RT, Bookout AL, He W, Straume M, Mangelsdorf DJ, Evans RM: Nuclear receptor expression links the circadian clock to metabolism. Cell. 2006, 126 (4): 801-810. 10.1016/j.cell.2006.06.050.View ArticlePubMedGoogle Scholar
- Bugge A, Feng D, Everett LJ, Briggs ER, Mullican SE, Wang F, Jager J, Lazar MA: Rev-erbalpha and Rev-erbbeta coordinately protect the circadian clock and normal metabolic function. Genes Dev. 2012, 26 (7): 657-667. 10.1101/gad.186858.112.View ArticlePubMed CentralPubMedGoogle Scholar
- Ahuja HS, Szanto A, Nagy L, Davies PJ: The retinoid X receptor and its ligands: versatile regulators of metabolic function, cell differentiation and cell death. J Biol Regul Homeost Agents. 2003, 17 (1): 29-45.PubMedGoogle Scholar
- Kung T, Murphy KA, White LA: The aryl hydrocarbon receptor (AhR) pathway as a regulatory pathway for cell adhesion and matrix metabolism. Biochem Pharmacol. 2009, 77 (4): 536-546. 10.1016/j.bcp.2008.09.031.View ArticlePubMed CentralPubMedGoogle Scholar
- Cavadini G, Petrzilka S, Kohler P, Jud C, Tobler I, Birchler T, Fontana A: TNF-alpha suppresses the expression of clock genes by interfering with E-box-mediated transcription. Proc Natl Acad Sci. 2007, 104 (31): 12843-12848. 10.1073/pnas.0701466104.View ArticlePubMed CentralPubMedGoogle Scholar
- Giebultowicz JM: Molecular mechanism and cellular distribution of insect circadian clocks. Annu Rev Entomol. 2000, 45 (1): 769-793. 10.1146/annurev.ento.45.1.769.View ArticlePubMedGoogle Scholar
- Saper CB, Lu J, Chou TC, Gooley J: The hypothalamic integrator for circadian rhythms. Trends Neurosci. 2005, 28 (3): 152-157. 10.1016/j.tins.2004.12.009.View ArticlePubMedGoogle Scholar
- Steiger JL, Russek SJ: GABAA receptors: building the bridge between subunit mRNAs, their promoters, and cognate transcription factors. Pharmacol Ther. 2004, 101 (3): 259-281. 10.1016/j.pharmthera.2003.12.002.View ArticlePubMedGoogle Scholar
- Balsalobre A, Brown SA, Marcacci L, Tronche F, Kellendonk C, Reichardt HM, Schütz G, Schibler U: Resetting of circadian time in peripheral tissues by glucocorticoid signaling. Science. 2000, 289 (5488): 2344-2347. 10.1126/science.289.5488.2344.View ArticlePubMedGoogle Scholar
- Shioda S, Takenoya F, Yagi M, Wang L, Hori Y, Kageyama H: Neural networks of several novel neuropeptides involved in feeding regulation. Nutrition. 2008, 24 (9): 848-853. 10.1016/j.nut.2008.06.016.View ArticlePubMedGoogle Scholar
- Impey S, Obrietan K, Wong ST, Poser S, Yano S, Wayman G, Deloulme JC, Chan G, Storm DR: Cross talk between ERK and PKA is required for Ca2+ stimulation of CREB-dependent transcription and ERK nuclear translocation. Neuron. 1998, 21 (4): 869-883. 10.1016/S0896-6273(00)80602-9.View ArticlePubMedGoogle Scholar
- Ginty DD, Kornhauser JM, Thompson MA, Bading H, Mayo KE, Takahashi JS, Greenberg ME: Regulation of CREB phosphorylation in the suprachiasmatic nucleus by light and a circadian clock. Science. 1993, 260 (5105): 238-241. 10.1126/science.8097062.View ArticlePubMedGoogle Scholar
- Wright KP, Gronfier C, Duffy JF, Czeisler CA: Intrinsic period and light intensity determine the phase relationship between melatonin and sleep in humans. J Biol Rhythm. 2005, 20 (2): 168-177. 10.1177/0748730404274265.View ArticleGoogle Scholar
- Buhr ED, Yoo S-H, Takahashi JS: Temperature as a universal resetting Cue for mammalian circadian oscillators. Science. 2010, 330 (6002): 379-385. 10.1126/science.1195262.View ArticlePubMed CentralPubMedGoogle Scholar
- Yang Q, Orman MA, Berthiaume F, Ierapetritou MG, Androulakis IP: Dynamics of short-term gene expression profiling in liver following thermal injury. J Surg Res. 2012, 176 (2): 549-558. 10.1016/j.jss.2011.09.052.View ArticlePubMed CentralPubMedGoogle Scholar
- Li C, Wong WH: Model-based analysis of oligonucleotide arrays: expression index computation and outlier detection. Proc Natl Acad Sci U S A. 2001, 98 (1): 31-36. 10.1073/pnas.98.1.31.View ArticlePubMed CentralPubMedGoogle Scholar
- Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, Ellis B, Gautier L, Ge Y, Gentry J, Hornik K, Hothorn T, Huber W, Iacus S, Irizarry R, Leisch F, Li C, Maechler M, Rossini AJ, Sawitzki G, Smith C, Smyth G, Tierney L, Yang JY, Zhang J: Bioconductor: open software development for computational biology and bioinformatics. Genome Biol. 2004, 5 (10): R80-10.1186/gb-2004-5-10-r80.View ArticlePubMed CentralPubMedGoogle Scholar
- Storey JD: The positive false discovery rate: A Bayesian interpretation and the q-value. Ann Stat. 2003, 31 (6): 2013-2035. 10.1214/aos/1074290335.View ArticleGoogle Scholar
- Tong W, Cao X, Harris S, Sun H, Fang H, Fuscoe J, Harris A, Hong H, Xie Q, Perkins R: ArrayTrack–supporting toxicogenomic research at the US Food and Drug Administration National Center for Toxicological Research. Environ Health Perspect. 2003, 111 (15): 1819-1826. 10.1289/ehp.6497.View ArticlePubMed CentralPubMedGoogle Scholar
- Genomatix. http://www.genomatix.de,
- Nguyen TT, Foteinou PT, Calvano SE, Lowry SF, Androulakis IP: Computational identification of transcriptional regulators in human endotoxemia. PLoS One. 2011, 6 (5): e18889-10.1371/journal.pone.0018889.View ArticlePubMed CentralPubMedGoogle Scholar
- Cartharius K, Frech K, Grote K, Klocke B, Haltmeier M, Klingenhoff A, Frisch M, Bayerlein M, Werner T: MatInspector and beyond: promoter analysis based on transcription factor binding sites. Bioinformatics. 2005, 21 (13): 2933-2942. 10.1093/bioinformatics/bti473.View ArticlePubMedGoogle Scholar
- Nguyen TT, Almon RR, Dubois DC, Jusko WJ, Androulakis IP: Comparative analysis of acute and chronic corticosteroid pharmacogenomic effects in rat liver: transcriptional dynamics and regulatory structures. BMC Bioinforma. 2010, 11: 515-10.1186/1471-2105-11-515.View ArticleGoogle Scholar
- Morgenstern B: DIALIGN 2: improvement of the segment-to-segment approach to multiple sequence alignment. Bioinformatics. 1999, 15 (3): 211-218. 10.1093/bioinformatics/15.3.211.View ArticlePubMedGoogle Scholar
- Sukumaran S, Almon RR, DuBois DC, Jusko WJ: Circadian rhythms in gene expression: Relationship to physiology, disease, drug disposition and drug action. Adv Drug Deliv Rev. 2010, 62 (9–10): 904-917.View ArticlePubMed CentralPubMedGoogle Scholar
- Bozek K, Relogio A, Kielbasa SM, Heine M, Dame C, Kramer A, Herzel H: Regulation of clock-controlled genes in mammals. PLoS One. 2009, 4 (3): e4882-10.1371/journal.pone.0004882.View ArticlePubMed CentralPubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.