Multilayer perceptron tumour diagnosis based on chromatography analysis of urinary nucleosides

Development and Assessment of Assisted Diagnosis Models Using Machine Learning for Identifying Elderly Patients With Malnutrition: Cohort Study

BACKGROUND

Older patients are at an increased risk of malnutrition due to many factors related to poor clinical outcomes.

OBJECTIVE

This study aims to develop an assisted diagnosis model using machine learning (ML) for identifying older patients with malnutrition and providing the focus of individualized treatment.

METHODS

We reanalyzed a multicenter, observational cohort study including 2660 older patients. Baseline malnutrition was defined using the global leadership initiative on malnutrition (GLIM) criteria, and the study population was randomly divided into a derivation group (2128/2660, 80%) and a validation group (532/2660, 20%). We applied 5 ML algorithms and further explored the relationship between features and the risk of malnutrition by using the Shapley additive explanations visualization method.

RESULTS

The proposed ML models were capable to identify older patients with malnutrition. In the external validation cohort, the top 3 models by the area under the receiver operating characteristic curve were light gradient boosting machine (92.1%), extreme gradient boosting (91.9%), and the random forest model (91.5%). Additionally, the analysis of the importance of features revealed that BMI, weight loss, and calf circumference were the strongest predictors to affect GLIM. A BMI of below 21 kg/m2 was associated with a higher risk of GLIM in older people.

CONCLUSIONS

We developed ML models for assisting diagnosis of malnutrition based on the GLIM criteria. The cutoff values of laboratory tests generated by Shapley additive explanations could provide references for the identification of malnutrition.

TRIAL REGISTRATION

Chinese Clinical Trial Registry ChiCTR-EPC-14005253; https://www.chictr.org.cn/showproj.aspx?proj=9542.

Quantifying RNA modifications by mass spectrometry: a novel source of biomarkers in oncology

Despite significant progress in targeted therapies, cancer recurrence remains a major cause of mortality worldwide. Identification of accurate biomarkers, through molecular profiling in healthy and cancer patient samples, will improve diagnosis and promote personalized medicine. While genetic and epigenetic alterations of DNA are currently exploited as cancer biomarkers, their robustness is limited by tumor heterogeneity. Recently, cancer-associated changes in RNA marks have emerged as a promising source of diagnostic and prognostic biomarkers. RNA epigenetics (also known as epitranscriptomics) is an emerging field in which at least 150 chemical modifications in all types of RNA (mRNA, tRNA, lncRNA, rRNA, and microRNA) have been detected. These modifications fine-tune gene expression in both physiological and pathological processes. A growing number of studies have established links between specific modified nucleoside levels in solid/liquid biopsies, and cancer onset and progression. In this review, we highlight the potential role of epitranscriptomic markers in refining cancer diagnosis and/or prognosis. RNA modification patterns may contain important information for establishing an initial diagnosis, monitoring disease evolution, and predicting response to treatment. Furthermore, recent developments in mass spectrometry allow reliable quantification of RNA marks in solid biopsies and biological fluids. We discuss the great potential of mass spectrometry for identifying epitranscriptomic biomarker signatures in cancer diagnosis. While there are various methods to quantify modified nucleosides, most are unable to detect and quantify more than one type of RNA modification at a time. Mass spectrometry analyses, especially GC-MS/MS and LC-MS/MS, overcome this limitation and simultaneously detect modified nucleosides by multiple reaction monitoring. Indeed, several groups are currently validating mass spectrometry methods that quantify several nucleosides at one time in liquid biopsies. The challenge now is to exploit these powerful analytical tools to establish epitranscriptomic signatures that should open new perspectives in personalized medicine. This review summarizes the growing clinical field of analysis of RNA modifications and discusses pre-analytical and analytical approaches, focusing in particular on the development of new mass spectrometry tools and their clinical applications.

Targeted metabolomics: Liquid chromatography coupled to mass spectrometry method development and validation for the identification and quantitation of modified nucleosides as putative cancer biomarkers

A notable change in the body fluids nucleosides of cancer patients has been actively highlighted in searches for new biomarkers to early cancer detection. For this reason, improvements of bioanalytical methods for these compounds focused on a noninvasive sampling trend are of great importance. Therefore, this work aimed firstly to develop efficient methods for nucleoside analysis in urine and serum by liquid chromatography-tandem mass spectrometry (LC-MS/MS), applying different strategies to quantify nine nucleosides, and further identify other untargeted nucleosides. Sample preparation was based on protein precipitation and affinity-solid phase extraction (SPE), whereas quantification was performed using a triple quadrupole (QqQ) mass analyzer operating in the selected reaction monitoring (SRM) mode. Surrogates matrices were proposed as an alternative to standard addition calibration. Specifically, to quantitate creatinine, a simple LC-MS/MS method was validated and used for normalization of urinary metabolites quantitation. To identify the other nucleosides, LC methods using different MS scans modes were evaluated on a quadrupole-time of flight (Q-TOF) or a hybrid triple quadrupole linear ion trap (Q-trap). Validation was performed for nucleosides quantification using the synthetic matrices of urine and serum, and selectivity, linearity, accuracy, reproducibility, matrix effect, LOD's and LOQ's were accessed, providing trustworthy results for bioanalysis purposes. Both LC-Q-Trap/MS and LC-Q-TOF/MS methods showed proper sensitivity for structural characterization on assays with urine and serum samples from healthy volunteers and could also be used in the identification of untargeted nucleosides. The investigated approaches delivered in-depth results and seem promising for future applications on urine and serum samples analyses aiming to validate nucleosides as cancer biomarkers.

Modified nucleosides as biomarkers for early cancer diagnose in exposed populations

There is increasing worldwide interest in developing of markers for tumor diagnosis and identification of individuals who are at high cancer risk. Cancer, like other diseases accompanied by metabolic disorders, causes characteristic effects on cell turnover rate, activity of modifying enzymes, and RNA/DNA modifications. This results in an increased excretion of modified nucleosides in cancer patients. Therefore, for many years modified nucleosides have been suggested as tumor markers. The aim of the study was to elucidate further the usefulness of urinary nucleosides as possible markers at early detection of cancer in persons which are exposed against tumor promoting influences during their working life. Uranium miners are exposed to many kinds of pollutants that can cause health damage even lead to carcinogenesis. We analyzed modified nucleosides in urine samples from 92 miners who are at high risk for lung cancer to assess the levels of nucleosides by a multilayer perceptron (MLP) classifier - a neural network model. Eighteen nucleosides/metabolites were detected with reversed-phase high-pressure liquid chromatography (RP-HPLC). A valid set of urinary metabolites were selected and multivariate statistical technique of multilayer perceptron neural network were applied. In a previous study, MLP shows a sensitivity and specificity of 97 and 85%, respectively. MLP classification including the most relevant markers/nucleosides clearly demonstrates the elevation of RNA metabolism in miners, which is associated with possible malignant disease. We found that there were 30 subjects with early health disorders among 92 uranium workers based on MLP technique using modified nucleosides. The combination of RP-HPLC analysis of modified nucleosides and subsequent MLP analyses represents a promising tool for the development of a non-invasive prediction system and may assist in developing management and surveillance procedures.

Statistical-based approach in potential diagnostic application of urinary nucleosides in urogenital tract cancer

Aim: We aimed at evaluation the potential diagnostic role of urinary nucleosides in urogenital tract cancer. Materials & methods: Concentrations of 12 nucleosides determined by LC-MS/MS were subjected to correlation, association and interaction analyses. Results: We identified six pairs of nucleosides differently correlated in the group of patients and controls (p < 0.05). N-2-methylguanosine (odds ratio: 4.82; 95% CI: 1.78–12.93; p = 0.002) and N,N-dimethylguanosine (odds ratio: 5.45; 95% CI: 1.78–16.44; p = 0.003), were significantly associated with the disease risk (p-corrected = 0.004). Interaction between N-2-methylguanosine and adenosine (p-interaction = 0.019) suggested their multiplicative effect on the outcome. Conclusion: Urinary nucleosides, namely N,N-dimethylguanosine and N-2-methylguanosine may have the potential to serve as prognostic biomarkers. Gender-specific differences in urogenital tract cancer are likely to occur.

A novel method of liquid chromatography-tandem mass spectrometry combined with chemical derivatization for the determination of ribonucleosides in urine

Ribonucleosides are the end products of RNA metabolism. These metabolites, especially the modified ribonucleosides, have been extensively evaluated as cancer-related biomarkers. However, the determination of urinary ribonucleosides is still a challenge due to their low abundance, high polarity and serious matrix interferences in urine samples. In this study, a derivatization method based on a chemical reaction between ribonucleosides and acetone to form acetonides was developed for the determination of urinary ribonucleosides. The derivative products, acetonides, were detected by using liquid chromatography-tandem mass spectrometry (LC-MS/MS). The methodological evaluation was performed by quantifying four nucleosides for linear range, average recovery, precision, accuracy and stability. The validated procedures were applied to screen modified ribonucleosides in urine samples. Improvement of separation and enhancement of sensitivity were obtained in the analysis. To identify ribonucleosides, inexpensive isotope labeling acetone (acetone-d6) and label-free acetone were applied to form ordinary and deuterated acetonides, respectively. The two groups of samples were separated with orthogonal partial least squares (OPLS). The ordinary and deuterated pairs of acetonides were symmetrically distributed in the S-plot for easy and visual signal identification. After structural confirmation, a total of 56 ribonucleosides were detected, 52 of which were modified ribonucleosides. The application of derivatization, deuterium-labeling and multivariate statistical analysis offers a new option for selective detection of ribonucleosides in biological samples.

The state-of-the-art determination of urinary nucleosides using chromatographic techniques "hyphenated" with advanced bioinformatic methods

Over the last decade metabolomics has gained increasing popularity and significance in life sciences. Together with genomics, transcriptomics and proteomics, metabolomics provides additional information on specific reactions occurring in humans, allowing us to understand some of the metabolic pathways in pathological processes. Abnormal levels of such metabolites as nucleosides in the urine of cancer patients (abnormal in relation to the levels observed in healthy volunteers) seem to be an original potential diagnostic marker of carcinogenesis. However, the expectations regarding the diagnostic value of nucleosides may only be justified once an appropriate analytical procedure has been applied for their determination. The achievement of good specificity, sensitivity and reproducibility of the analysis depends on the right choice of the phases (e.g. sample pretreatment procedure), the analytical technique and the bioinformatic approach. Improving the techniques and methods applied implies greater interest in exploration of reliable diagnostic markers. This review covers the last 11 years of determination of urinary nucleosides conducted with the use of high-performance liquid chromatography in conjunction with various types of detection, sample pretreatment methods as well as bioinformatic data processing procedures.

A fast capillary electrophoresis method for separation and quantification of modified nucleosides in urinary samples

Modified nucleosides are formed at the post-transcriptional stage by chemical modification of normal nucleosides within the ribonucleic acid (RNA). These modified nucleosides cannot be reutilized or further degraded, but they are excreted in the urine as intact molecules. The elevated levels of modified nucleosides in the urine samples have served as potential cancer biomarkers in many studies. Although different analytical techniques have been reported for determining nucleosides levels, they are practically difficult to use as a routine tool for early cancer screening. In this paper, a novel method was developed to separate and quantify 10 nucleosides--adenosine, cytidine, guanosine, uridine, inosine, xanthosine, pseudouridine, N(2)-methylguanosine, 1-methyladenosine, and N(2),N(2)-dimethylguanosine--in urine samples using capillary electrophoresis with an ultraviolet (UV) detector (abbreviated as CE-UV) at a wavelength of 254 nm. A 50 microm (i.d.) x 38 cm (effective length) fused silica capillary was used for the separation, and a borate-phosphate buffer containing 25 mM cetyltrimethylammonium bromide (CTAB) at pH 9.50 was used as a background electrolyte. The separation was performed at 15 kV under reverse polarity and completed within 10 min. The linear range of the analytes was 5.0-500 micromol/L, and the limit of detection was <2.0 micromol/L. The effects of pH, buffer concentrations, CTAB concentration, and the operation voltages on the separation and quantification of the modified nucleosides were also investigated. The technique developed in this study is much simpler and faster, compared to previous studies, and can be used to quantify modified nucleosides in urine samples.

Analysis of urinary nucleosides as potential tumor markers in human colorectal cancer by high performance liquid chromatography/electrospray ionization tandem mass spectrometry

BACKGROUND

Increased levels of modified nucleosides have been observed in urine from patients suffering from several cancers. In this study, we evaluated whether urinary nucleosides can serve as potential tumor markers for colorectal cancer by high performance liquid chromatography-electrospray/tandem mass spectrometry (HPLC/ESI-MS/MS).

METHODS

A simple and specific method based on HPLC/ESI-MS/MS was developed to determine the urinary nucleosides from patients with colorectal cancer. We studied the excretion patterns of nucleosides in urine from 26 patients with colorectal cancer and 18 healthy controls.

RESULTS

The LC/MS/MS approach with selective reaction monitoring (SRM) allowed for the sensitive determination of nucleosides in human urine samples with colorectal cancer. The mean levels of 5 urinary nucleosides (adenosine, cytidine, N(2),N(2)-dimethylguanine, 8-hydroxy-2'-deoxyguanosine and uridine) were significantly higher in the patients with colorectal cancer than in the healthy adults.

CONCLUSIONS

The results indicate that urinary nucleosides determined by LC/MS/MS may be useful as biological markers for colorectal cancer. Our findings suggest that LC/MS/MS is a highly specific and sensitive method for rapidly screening a large number of nucleoside that may be useful as markers for cancer in humans.