Protocol for the isolation, sequencing, and analysis of the gut phageome from human fecal samples

Summary The phage-bacteria interactions in the gut microbiome are critical for health and disease, but viruses of the human gut microbiome are poorly understood. Here, we present a simple and cost-efficient protocol for collecting viral-like particles (VLPs) from human fecal samples. We describe VLPs quantification using epifluorescence and TEM microscopy, followed by DNA sequencing and bioinformatics analysis. This protocol characterizes the gut phageome in normal-weight and obese children with metabolic syndrome. It is also suitable to conduct high-throughput studies for other diseases. For complete details on the use and execution of this profile, please refer to Bikel et al. (2021).


MATERIALS AND EQUIPMENT
The essential equipment of a molecular microbiology laboratory is also needed, including a conventional benchtop microcentrifuge, a vortex mixer, an electrophoresis chamber, and a laminar airflow cabinet. Also, expendable materials, including pipette tips, microcentrifuge, and PCR tubes, are necessary. In this protocol, the Applied Biosystems SimpliAmpä Thermal Cycler was used for the library index PCR.

Viral-like particles (VLPs) isolation
Timing: 6-7 h per sample This step is helpful to isolate and concentrate the VLPs from the fecal samples.
Before beginning the experiments, thaw the fecal samples on ice. Also, it is essential to filter all reagents through sterile syringe filter polyethersulfone membrane (PES) units. We recommend using a 0.45 mm pore size to remove larger cell debris followed by a 0.22 mm pore size to remove remaining cell debris and bacterial-size particles. This two-step filtration protocol applies to all reagents and the supernatant of the fecal samples.
1. Preparation of phage suspension (15 min per sample). a. Pour 40 mL of SM Buffer, previously filtered, into a 50 mL centrifuge tube. b. Gently pipet the 250 mg of fecal sample and resuspend in RNA later several times until obtaining a homogeneous mixture. c. Add the whole RNA later + fecal sample mixture to the SM Buffer tube.
a. Centrifuge at 4700 g for 30 min at 4 C. b. Carefully transfer the supernatant (it contains the VLPs) to a 5 mL syringe attached to a 0.45 mm PES filter, using a micropipette. Filter the supernatant and collect the filtrate into a new 50 mL centrifuge tube. Note: When taking the supernatant with the micropipette, make sure not to disturb the pellet. We recommend using a fine tip and removing the volume very slowly.
c. Filter all the supernatant obtained from the previous step through a 0.22 mm PES filter.
Note: Regularly change the filters to avoid clogging them with cell debris that was not removed by the centrifugation step.
Note: Do not push too hard with the syringe to avoid breaking the filter.
Pause Point: The filtered buffer with VLPs can be stored at 4 C until needed.
Note: Some reports have shown good stability of phages when stored at 4 C in SM buffer for more than six months (Jepson and March, 2004;Jo nczyk et al., 2011). Similarly, storing at 2 C-5 C has shown no significant reduction in phage titer (5%-10%) after six months (Mullan 2001).
Note: Check that the SM buffer contains gelatin. It stabilizes VLPs membranes during storage.
3. Enrichment and washing the VLPs (2-3 h). a. Add 15 mL of the previous buffer with VLPs into an Amicon Ultra-15 centrifugal filter (100 KDa). b. Centrifuge at 5000 g in a fixed-angle centrifuge for 15-60 min at 4 C, until the remaining volume reaches $200 mL on the top of the Amicon filter unit. c. Discard the flow-through and add fresh filtrate again to the top of the filter unit. d. Repeat steps a, b and c until the entire buffer is filtered. e. Add 5 mL of SM Buffer to the top chamber of the Amicon and spin at 5000 g at 4 C for 2-5 min to wash the VLPs until the remaining volume reaches $200 uL on the top of the Amicon filter unit. Repeat this VLPs washing step (at least five times) until the buffer containing the VLPs is clear.
Note: Wash several times the VLPs to diminish cellular debris that could inhibit the nucleic acid extraction and their visualization in the microscope. Eight washes of the VLPs allow a more precise visualization of viral particles by electron microscopy (TEM).
f. Discard the collector tube of the Amicon filter unit and pipet the $200 mL of VLPs several times to resuspend the VLPs retained on the filter walls. g. Transfer the total volume $200 mL containing the VLPs to a 1.5 mL microfuge tube.
Note: If the volume in the Amicon filter unit is less than 200 mL, adjust the volume using SM buffer. Note: Wash two times the VLPs to clean up remnants of the DNAse reaction.
d. Transfer the previous reaction into an Amicon Ultra-15 centrifugal filter (100 KDa) and add 15 mL of SM buffer. e. Centrifuge at 5000 g in a fixed-angle centrifuge for 15-60 min at 4 C, until the remaining volume reaches $200 uL on the top of the Amicon filter unit. f. Discard the flow-through, add fresh SM buffer to the top of the filter unit, and repeat step e. g. Discard the collector tube of the Amicon filter unit and pipet the $200 mL of VLPs several times to resuspend the VLPs retained on the filter walls. h. Transfer the total volume $200 mL containing the VLPs to a 1.5 mL microfuge tube.
Pause Point: The VLPs can be stored at 4 C until needed.
Note: At this point it is recommended to use TEM microscopy to confirm the presence of VLPs and count the number of particles using fluorescence microscopy.

Microscopy visualization and VLPs counts
Timing: 2-2.5 h per sample This step is to corroborate the presence and quantify the VLPs in the samples.
a. Preparation of the SYBR Green staining solution (5 min). i. Dilute 5 mL of SYBR Green in 45 mL of nuclease-free water (previously passed through the 0.22 mm PES filtration units). b. Staining the VLPs of the sample (10 min).
i. Place 10 mL of the VLPs sample in a 0.6 mL Eppendorf tube.
ii. Add 2 mL of the SYBR Green solution.
CRITICAL: Paraformaldehyde is a toxic chemical, and it only should be used in a chemical fume hood. Waste should be collected and discarded appropriately.
c. Mounting the VLPs sample on a microscopy slide (5 min). i. Place 10 mL of stained VLPs sample on a slide. ii. Place a coverslip over the sample, avoiding the formation of air bubbles. d. Microscopy visualization and quantification of VLPs (Olympus FV1000 Multi-photonic confocal microscope) (1h). i. Place the slide under the microscope and focus on the field at the confocal microscope. Set the laser's parameters to the optimal setup. ii. Take five micrographs of the sample, each in triplicate. This protocol uses 0.212 3 0.212millimeter images ( Figure 1A). iii. Use the FIJI software (Schindelin et al., 2012)  field, taking into account the area of one field, and back calculating based on the dilution factor, as follows:  Note: The optimal microscope setup depends on the microscope type and operating system.
Note: Fluorescence is not stable over time. Thus the images should be taken immediately after slide preparation.
Note: Some dyes can stain the phage capsid rather than staining the DNA inside it. Both stains can be used depending on the experimental need. Keep in mind that the SYBR stain does not modify the exterior of the phage.
7. Corroborate the phage morphology using Transmission Electron Microscopy (TEM) (1 h). a. Place 8 mL of a concentrated VLP sample onto carbon-coated Formvar grids. b. Incubate at room temperature for 1 min. c. Drain off the sample excess of the grid with the aid of filter paper. d. Stain the grids with 8 mL of 2% uranyl acetate. e. Incubate at room temperature for 2-3 min. f. Drain off the excess stain with the aid of filter paper. g. Dry the grids at room temperature until analysis. h. Take the image from the grids at magnifications of 14,0003 to 80,0003 ( Figure 1B).

CRITICAL:
The image fields could be obstructed in a few instances by extensive, amorphous, dark-staining material. (See potential problem 1 in the troubleshooting section). Note: Storage at -20 C will prolong its life, but avoid repeated freezing and thawing. Dividing the solution into aliquots and freezing at -20 C is recommended.

Viral DNA extraction and purification
9. Preparation of Buffer AW1. a. Add 25 mL of ethanol (96%-100%) to a bottle containing 19 mL of Buffer AW1 concentrate, as described on the bottle. b. Tick the check box on the label to indicate that ethanol has been added. c. Store reconstituted Buffer AW1 at room temperature (15 C-25 C).
Note: Reconstituted Buffer AW1 is stable for up to 1 year or until the kit expiration date when stored at room temperature.
Note: Always mix reconstituted Buffer AW1 by shaking before starting the procedure.
CRITICAL: Buffer AW1 contains chaotropic salt. Take appropriate laboratory safety measures and wear gloves when handling. Not compatible with disinfectants containing bleach.

Preparation of Buffer AW2
a. Add 30 mL of ethanol (96%-100%) to a bottle containing 13 mL of Buffer AW2 concentrate, as described on the bottle. b. Tick the check box on the label to indicate that ethanol has been added. c. Store reconstituted Buffer AW2 at room temperature (15 C-25 C).
Note: Reconstituted Buffer AW2 is stable for up to 1 year or until the kit expiration date when stored at room temperature.
Note: Always mix reconstituted Buffer AW2 by shaking before starting the procedure.

Viral DNA extraction
Before starting, let the samples reach room temperature. All centrifugation steps were carried out at room temperature (15 C-25 C). a. Pipet 25 mL QIAGEN Protease into a 1.5 mL microcentrifuge tube b. Add the 200 mL of the VLPs sample into the microcentrifuge tube.
Note: If the sample volume is less than 200 mL, add the appropriate volume of SM Buffer solution to bring the volume of the protease and the sample to a total of 225 mL. c. Briefly centrifuge the 1.5 mL tube to remove drops from the inside of the lid. d. Carefully apply all lysates onto the QIAamp MinElute column without wetting the rim. Close the cap and centrifuge at 60003g (8000 rpm) for 1 min. e. Place the QIAamp MinElute column in a clean 2 mL collection tube, and discard the collection tube containing the filtrate.
Note: If the lysate has not entirely passed through the column after centrifugation, centrifuge again at a higher speed until the QIAamp MinElute column is empty.
f. Carefully open the QIAamp MinElute column, and add 500 mL of Buffer AW1 without wetting the rim. g. Close the cap and centrifuge at 60003g (8000 rpm) for 1 min. h. Place the QIAamp MinElute column in a clean 2 mL collection tube, and discard the collection tube containing the filtrate. i. Carefully open the QIAamp MinElute column, and add 500 mL of Buffer AW2 without wetting the rim. j. Close the cap and centrifuge at 60003g (8000 rpm) for 1 min. k. Place the QIAamp MinElute column in a clean 2 mL collection tube, and discard the collection tube containing the filtrate. l. Carefully open the QIAamp MinElute column and add 500 mL of ethanol (96-100%) without wetting the rim. m. Close the cap and centrifuge at 60003g (8000 rpm) for 1 min. n. Discard the collection tube containing the filtrate. o. Place the QIAamp MinElute column in a clean 2 mL collection tube. Centrifuge at full speed (20,0003g; 14,000 rpm) for 3 min to dry the membrane completely.
Note: To evaporate any remaining liquid, it is recommended to place the QIAamp MinElute column into a new 2 mL collection tube (not provided), open the lid, and incubate the assembly at 56 C for 3 min to dry the membrane completely.
p. Place the QIAamp MinElute column in a clean 1.5 mL microcentrifuge tube, and discard the collection tube with the filtrate. q. Carefully open the lid of the QIAamp MinElute column, and apply 30 mL of Buffer AVE or RNase-free water to the center of the membrane. r. Close the lid and incubate at room temperature for 1 min.
Note: Incubating the QIAamp MinElute column loaded with Buffer AVE or water for 5 min at room temperature before centrifugation generally increases DNA yield.
s. Centrifuge at full speed (20,0003g; 14,000 rpm) for 1 min. t. Eluted DNA can be collected in standard 1.5 mL microcentrifuge tubes. Note: Ensure that the elution buffer is at room temperature. If elution is done in small volumes (<50 mL), the elution buffer must be dispensed onto the center of the membrane for complete elution of bound DNA.
Note: Elution volume is flexible and can be adapted according to the requirements of the downstream application. However, the recovered elution volume will be approximately 5 mL less than the volume of elution buffer applied onto the column.
Note: If the purified viral DNA is used within 24 h, storage at 2 C-8 C. For periods longer than 24 h, storage at -20 C.
Note: With the application of this protocol, we obtained an average of 160.7 G 105.0 ng of total DNA from 200 mL of VLPs extracted from 250 mg of feces (Bikel et al., 2021).

DNA library preparation and purification
Timing: 1.5-2 h per sample In this step, the sequencing libraries of the extracted DNA from the VLPs are prepared using the Nextera XT DNA Library Preparation Guide (https://support.illumina.com/content/dam/illuminasupport/documents/documentation/chemistry_documentation/samplepreps_nextera/nextera-xt/ nextera-xt-library-prep-reference-guide-15031942-05.pdf). Prepare the sequencing libraries in a PCR laminar airflow cabinet to avoid contamination.
a. Use an electrophoresis gel and a UV absorbance method to assess the quality of the DNA sample. Absorbance ratio values of 1.8-2.0 are considered adequate for this protocol. 14. Quantification of input DNA.
a. To accurately quantify the input DNA and library concentration, use a fluorometric-based method such as Qubit dsDNA Assay. b. Dilute the sample DNA to a final concentration of 0.3 ng/mL in nuclease-free water.
CRITICAL: Avoid using DNA quantification methods that measure total nucleic acid content, such as Nanodrop or other UV absorbance methods. Contaminants such as ssDNA, RNA, and oligonucleotides are not substrates for the Nextera XT assay.
CRITICAL: A DNA concentration of 0.3 ng/mL is essential because the Nextera XT Protocol is optimized for $1 ng of input DNA contained in 5 mL. If the total concentration of DNA extracted from the VLP samples results low, an additional DNA concentration step might be needed.
Note: Maintain the DNA sample on the ice during the process or store it at À20 C until use.
The Nextera XT transposase simultaneously fragments the input DNA and adds adapter sequences during this step. The reactions can be assembled in sterile PCR tubes. a. Thaw the diluted DNA sample and the following Kit reagents on ice.
ii. Maintain the Neutralize Tagment Buffer (NT) at room temperature. b. Vortex and visually inspect all reagents to make sure there is no precipitated. Suppose there is, vortex gently until the precipitate is resuspended.
Note: Assemble the reaction in the order described for optimal kit performance. The reaction does not need to be assembled on ice.

OPEN ACCESS
c. Add 10 mL TD buffer to a new PCR tube. d. Add 5 mL input DNA 0.3 ng/mL (1.5 ng total) to the tube with the TD buffer and gently pipette up and down five times to mix.
CRITICAL: Use the same amount of input DNA for all libraries.
Note: The user can assemble more than one library simultaneously. However, we recommend only preparing together libraries that belong to the same experimental group to avoid cross-contamination among tested groups.
e. Add 5 mL ATM buffer to the reaction tube containing the TD buffer and the DNA. Gently pipette up and down five times to mix. f. Centrifuge the reaction tube at 2803g at room temperature for 1 min. g. Incubate the reaction tube in a thermal cycler at 55 for 5 min and then hold at 10 C. Maintain the thermal cycler lid heated during the incubation. h. When the incubation reaches 10 C, immediately add 5 mL of the NT buffer to the reaction tube to neutralize the reaction. Pipette up and down five times to mix.
Note: To immediately neutralize the tagmentation reaction, adding the NT buffer while the reaction tube is still in the thermal cycler is more convenient.
i. Centrifuge the reaction tube at 2803g at room temperature for 1 min. j. Maintain the samples at room temperature for 5 min.
16. PCR Amplification. During this step, the tagmented DNA is amplified for a limited number of PCR cycles to add the index i7 and i5. a. Select a pair of one i7 plus one i5 index primer for each library.
CRITICAL: Two libraries cannot have the same pair of index primers, as they would be interpreted and sequenced as the same sample.
b. Thaw the selected primers and the Nextera PCR Master Mix (NPM) on ice. i. After all reagents are thawed mix each tube by inverting 3-5 times and centrifuge at 2803g for 1 min. c. Add 15 mL of NPM to the tagmented DNA sample. d. Add 5 mL of index i7 and 5 mL of index i5 to each sample. Gently pipette up and down five times to mix.

CRITICAL: Change tips between index primers to avoid cross-contamination.
e. Centrifuge at 2803g at room temperature for 1 min. f. Perform PCR using the following program and maintain the lid heated during the process: CRITICAL: Use the same number of PCR cycles for all libraries to avoid over-estimation in a particular library. Adding extra cycles of PCR could produce low-quality sequence results.
Pause Point: The user can safely stop before proceeding to PCR clean-up. The user can either keep the samples at 10 C overnight (12-14 h) inside the thermal cycler or store at 2 C-8 C for up to 2 days.

PCR Clean-Up.
This step uses AMPure XP beads to purify the DNA library and removes short library fragments.
Note: Before start, bring AMPure XP beads to room temperature.
a. Vortex the AMPure XP beads to evenly disperse them. b. Add 30 mL of AMPure XP beads to each amplified library. Gently pipette up and down ten times to mix.
Note: The volume of 30 mL of AMPure beads selects inserts of >500 bp. Refer to the Nextera XT DNA Library Preparation Guide to select the desired insert sizes.
c. Incubate at room temperature for 5 min. d. Place the tube on a magnetic stand until the supernatant has cleared. e. While the tube is on the magnetic stand, carefully remove and discard the supernatant.
Note: If any beads are aspirated, dispense back to the tube and wait until the supernatant is clear before removing again.
f. While the tube is on the magnetic stand, add 200 mL of 80% ethanol to wash the beads.
CRITICAL: Prepare fresh 80% ethanol from absolute ethanol. Ethanol can absorb water from the air altering the concentration and impacting the results.
g. Maintain the tube on the magnetic stand and incubate for 30 s. h. Remove and discard the supernatant carefully. i. Maintain the tube on the magnetic stand and allow the beads to air-dry for 15 min.
CRITICAL: Do not over-dry the beads. This could reduce the amount of library recovered.
j. Remove any excess ethanol with a clean tip. k. Remove the tube from the magnetic stand and add 50 mL of nuclease-free water. Gently pipette up and down ten times to mix. l. Incubate at room temperature for 2 min. m. Place the tube on the magnetic stand and wait until the supernatant has cleared. n. While the tube is on the magnetic stand, transfer the supernatant carefully to a clean tube. o. Quantify the final library with a fluorometric-based method such as Qubit dsDNA assay, and check the size distribution in a High Sensitivity DNA Bioanalyzer. p. All final libraries can be pooled together and sequenced in any Illumina platform.
Note: To achieve an even sequencing depth among all samples is essential that all libraries are pooled in equimolar proportions. As a reference, we adjusted each library to a final concentration of 2nM (Bikel et al., 2021).
Note: The Nextera libraries contain Illumina adapters; thus, they could be sequenced in any Illumina platform. As a reference, we used the NextSeq500 system in a 2 3 150 pair-end ll OPEN ACCESS mode and obtained 74,859,356 reads (an average of 2,673,548 reads per sample) (Bikel et al., 2021).
Note: According to Illumina's recommendations, the sequencing read length for De novo sequencing ranges from 2 3 150 to 2 3 300 bp. As a reference, when sequencing with 2 3 150 bp, we assembled 18,602 viral contigs with R500 nt of length (Bikel et al., 2021).

EXPECTED OUTCOMES
The library preparation with 1.5 ng of input DNA yields $200 ng of library with a size distribution >500 bp (Figure 2).

QUANTIFICATION AND STATISTICAL ANALYSIS
The data used for this protocol are in NCBI under BioProject accession number: PRJNA646512. In addition, all the scripts described below are in the Github repository https://github.com/lab8a/ 2021-iScience-Phageome.
Cleaning and clustering of sequenced reads 1. First, perform the quality visualization with FastQC.
CRITICAL: Prior to the beginning and after the pretreatment, visualize the quality of the reads with FastQC. The most important parameters are: 'Per base sequence quality', 'Per sequence content', 'Per base GC content', 'Per base N content', 'Overrepresented sequences' and 'Adapter content'. All of these parameters must be checked as Passed after the pretreatment.
4. Then, remove potential human and bacterial contamination. To achieve this, use the BWA aligner against the Homo Sapiens GRCh38.p13 reference genome (GenBank: GCA_000001405.28) and the Kraken database against bacteria NR database using the default parameters.
5. Finally, cluster the quality-filtered reads at 95% identity using CD-HIT to generate a unique and non-redundant dataset. Then, remove those reads mapped against human and bacterial genomes in the previous steps, the remaining set is named quality-filtered reads.
Note: The complete script for these steps is in the CommandLines_GLL-iScience.md file.
Analysis of viral read richness 6. Determine viral richness between groups by collecting 1,000 random subsamples of 149,000 single-end quality-filtered reads using seqtk subseq (the number of reads is determined by the sample with the minor sequencing depth). Perform this exercise to simulate a rarefaction analysis without including a taxonomic bias on the data.
>kraken -db krakenDB/bacteria_RefSeq -fastq-input -paired quality-reads-R1.fq quality- 7. Then, cluster each sub-sample at 95% identity with CD-HIT to identify the non-redundant and unique group of reads, and finally, count the reads per file.
Note: The complete script for this method can be found in the subseq_cdhit.md file.

Functional profiles and pVOGs analysis
8. Map quality filtered reads against the viral NR RefSeq and pVOGs databases using BLASTX with a maximum of 50 reported target sequences and a maximum e-value cutoff of 0.001. 9. Then, generate a relative abundance matrix using an in-house bash script. Annotate this matrix according to the KEGG classification of each predicted protein and the UniProtKB online database using an in-house bash script. For the Prokaryotic Virus Orthologous Groups (pVOGs), map the quality-filtered reads against this database with a maximum e-value cutoff of 0.001 and 50 reported target sequences. Finally, generate the matrix with the pVOGs classification per sample using an in-house bash script. All these in-house bash developed scripts are in the reads_blastx_pvogs.md file.
Classification of viral reads 11. Calculate the relative abundance per sample using the absolute read count for selected viral taxa and the total reads from each sample. This script is in the reads_blastx_megan.md file.
De novo contig assembly 12. Construct the de novo assembly using all the quality-filtered reads from all samples with the IDBA-UD assembler. However, due to a sample mixed-assembly, the chimeric contigs are filtrated only to obtain the contigs covering R80% of their total size by the viral reads in at least one sample.
13. Then, map each sample reads individually with Bowtie2 against the de novo viral assembly using the end-to-end mode with default parameters.
Note: Select only the viral scaffolds covered >80% in length by the reads of at least one sample to discard chimeras. Remove the scaffolds with less than 4 kB to have a higher chance of only keeping nearly complete viral genomes. Lastly, use CD-HIT with a 95% clustering identity to eliminate the redundant scaffolds. These scripts are in the CommandLines_GLL-iScience.md file. 16. Assign the final taxonomy of the BLASTX scaffolds with the LCA from MEGAN6, using the same parameters as described above for DC_MEGABLAST. These scripts are in the contigs_blastx_megan.md file. Finally, perform the VirSorter2 classification of each scaffold with the default parameters.
Note: The virsorter.md contains the steps of this analysis.
17. Perform the last classification using BLASTX against the Prokaryotic Virus Orthologous Groups (pVOGs) database with a maximum e-value cut-off of 0.001 and maximum target sequences to report set to 50.
Note: The script of this method can be consulted in the file reads_blastx_pvogs.md.

Differential abundance of phage contigs
18. Use the recruitment of reads to the contigs assembly to construct an abundance matrix. Define the coverage from reads mapping (Bowtie2) at R90% identity and R80% length. Convert the mapping outputs into a normalized abundance matrix using an in-house R script using the Reads Per Kilobase per Million sequenced reads per sample (RPKM). The name of this script is multi_-contingencty_table_transformations_taxa.R. The next line is an example of using this script.

Richness and diversity of phage contigs
19. Evaluate the contig assembly's richness and diversity based on the median of 10,000 rarefactions with a sequencing depth equal to the smallest sample based on the RPKM matrix. Conduct this process in QIIME 1.9. Assign the presence of phage contigs in the samples as either core phages: detected in >80% of the samples; common phages: in >50% and <80%; and individual phages: appearing in <50% of the population. These scripts are in the alpha_div_rarefaction.R and alpha_div_rar_compare.R files. Bacteria and biochemical parameters correlations 20. Calculate the correlations using the Spearman coefficient with rcorr function in R. Select the RPKM matrix for the contigs phage abundance and the relative frequency of the significant over-abundant taxa for O and OMS for the microbiota abundance. The script used for this step is in the spearman_corr.md file. The following is an example of use.

LIMITATIONS
A significant limitation for better understanding the role of the human gut virome in health and disease is the lack of standardized methods that allow high throughput virome analysis, coupled with the lack of a universal viral marker, unlike the 16S gene in the bacteriome. Therefore, studying the virome requires large-scale metagenomic sequencing approaches.
The first challenge in this study is the number of samples, which probably accounted for the lack of statistical significance obtained in some of the analyses.
We used the TAG method that strongly selects against ssDNA templates for library preparation, obtaining ssDNA viruses near or below detection limits. However, to eliminate the bias due to the TAG method, we decided to eliminate from the analysis all ssDNA viruses. Further improvements in sequencing library preparation techniques are essential to overcome this bias and bring metagenomic research of human gut phageome to a fully quantitative level.
The vast majority of contigs assembled from the data generated in this study could not be aligned to any known viral genomes in the NCBI RefSeq database, suggesting that many of the sequences are of unknown viral origin. In this regard, more virome studies are needed to gain a broader understanding of the composition of the human gut virome.
We performed an accurate estimation of viral community composition and diversity. However, we should note that our assemblies may represent fragments of the same phage genome that could affect accurate estimates about our phageome and its prevalence in the human population.

Problem 1
The filters clog very quickly with bacterial debris (step 2).

Potential solution
Increase the time used for centrifugation.

Problem 2
The filter was broken (step 2).

Potential solution
Change the filter for a new one and discard the old one.

Problem 3
Fluorescence is not observed in the microscope (step 6).

Potential solution
Prepare a new staining reaction and immediately observe after slide preparation.

Problem 4
In a few instances, view fields in TEM were obstructed by extensive, amorphous, dark-staining material (step 7).

Potential solution
One of the solutions for this problem is to increase the number of washes with the Amicon Ultra-15 Centrifugal Filter Units 100 KDa after all the filtrate is processed. To this end, add SM Buffer to the top chamber of the Amicon and spin at 5000 g for 2-5 min to wash the filtrate. Eight washes of the Amicon allow us a more precise visualization of viral particles by electron microscopy (TEM).

RESOURCE AVAILABILITY
Lead contact Further information and requests for resources and reagents should be directed to and fulfilled by the lead contact, Adrian Ochoa-Leyva (adrian.ochoa@ibt.unam.mx).

Materials availability
This study did not generate new unique reagents.
Data and code availability All original code have been deposited to GitHub: https://github.com/lab8a/2021-iScience-Phageome and in Zenodo: https://doi.org/10.5281/zenodo.5846703. The accession number for the sequenced data reported in this paper is NCBI BioProject: PRJNA646512. Accession numbers are also listed in the key resources table. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.