a. Sample Collection from Animals
Two commercial dairy farms nearby with a minimum of 50 animals were enrolled for the study. At both, farms animals were fed ad libitum with wheat hay, lucerne, and green grass. The concentrate mixture (cottonseed, groundnut, soybean, maize, etc) was fed at 2 kg per animal after boiling. Animals in both the farms were screened for brucella, TB, and JD before starting the experiment and were found negative. Sixteen multiparous cows with median 3 (1 to 5) lactations and 80 (70 to 95) days in milk were selected initially for the study. All cows were evaluated for reproductive health with rectal palpation and cervicovaginal discharge. Cows were administered 500 mcg cloprostenol (Pregma, Intas pharma, India) intramuscularly at 10-day intervals to promote lysis of corpus luteum and estrus synchronization. Using behavior signs, cervico-vaginal estrus discharge, rectal palpation, and synchronized estrus within 96 hours after second cloprostenol injection was considered inclusion criteria for final enrollment of cows. Out of 16, nine cows fulfilled the above were criteria and included in the study.
Four vaginal samples per cow were collected during the estrus (day 0: D0), metestrus (day 04: D4), diestrus (day 12: D12), and proestrus (day 16: D16) phase of estrous cycle. Before taking the vaginal samples, back-raking was performed followed by external genitalia were cleaned with 4% chlorhexidine and dried with tissue paper. Sterile vaginal Swab (17214/2950 Swab, Eppendorf Pvt ltd., India) were used to collect the vaginal samples from the fornix of the vagina. The samples were collected rotating swab towards clock viz direction for the 30s during four phases of the estrous cycle in cows. The samples were transported immediately in an icebox at 4 ℃ at the laboratory. Before processing, the samples were labeled according to the stage of the estrous cycle, cow number and farm identification.
Collection of blood samples on the respective days of the harmonized estrous cycle were day D0, D4, D12, and D16 from the jugular vein using 8 ml EDTA vacutainers. Blood was transported at 4℃ to the laboratory and plasma extracted after centrifuging at 4000 rpm (Thermo Scientific Sorvall X4R Pro-MD, India) for 5 minutes and separated plasma samples were labeled and stored in 2ml plasma storage vials at -20˚C until analysed for progesterone (P4) concentration by chemiluminisence immune assay (CLIA) method in a commercial laboratory (Gaievski et al 2018). Plasma P4 concentration < 1.0 ng/ml defined estrus and P4 concentration above > 1.0 ng/L defined other than estrus phase. The intra and inter assay coefficient of variation for P4 was 8.1 % and 8.6 %, respectively.
Nucleic acid Extraction and Amplification:
Vaginal swab samples from the nine cows were collected during the Estrous cycle at four different points (D0, D4, D12, and D16) for DNA isolation. DNA isolation from the swab sample was done using Qiagen Stool DNA isolation kit (cat. No. 51504) using the protocol provided in the manual of the kit. DNA quantification and purity were checked using the QIAxpert System. DNA was run on 0.8% agarose gel electrophoresis to check DNA integrity.
DNA was amplified using a specific primer of V3-V4 region of bacterial 16s rRNA gene Barcoded Fusion Forward Primer and Reverse Primer in Table 3 using thermal cycle condition. Denaturation: 1 min at 95 ºC, Annealing: 30 Sec at 58ºC, Extension: 45 secs at 75 ºC (30cycles). PCR product was run on 2% agarose gel electrophoresis to confirm the product size.
Library Preparation and Sequencing:
From the amplified product some of the products of the 16s rRNA gene showed nonspecific size amplification. Size-specific product (~450bp) purification was done using the E-Gel CloneWell II agarose gels. E-Gel CloneWell II agarose gel contains two comb systems. The sample was loaded into the upper well and electrophorese until the desired band came into the lower well. Then just simply the band is pipetted out and quantified using Qubit dsDNA HS Assay Kit with Qubit 2.0 Fluorometer. Further, remaining PCR products are purified using AMPure XP beads with a 0.9X ratio according to the user guide. After purification, Quantification was done using Qubit dsDNA HS Assay Kit with Qubit 4.0 Fluorometer before pooling of library and then sequencing was done using ion 5s platform using 530chip, with 400bp chemistry.
Amplicon Sequence analysis:
Raw reads were processed using a Perl script followed by prinseq lite. Data was filtered taking the quality value 25 and trimming of smaller sequences lower than 200bp. After filtration, the data were analysed using the software QIIME 2-2020.8  with default parameters until stated otherwise. DADA2 was used for the denoising and demultiplexing of the reads. The taxonomical analysis was done using the SILVA database . OTU clustering was done using 99% identity. Jaccrad, Bary Cutris, unweighted and weighted Unifrac distance were calculated to observe the difference in the community.
Alpha Rarefaction curve was generated using different metrics Shannon, observed feature, and Faith_Pd. Alpha diversity analysis was done using Kruskal-Wallis (pairwise) statistics. Beta diversity between the groups was calculated using permanova pairwise statistical analysis. % Relative abundance was calculated from taxa generated from the QIIME 2.0. The Taxonomic data were further analysed using the STAMP software  and statistical analysis was carried out using Multiple test ANOVA with log tra. The statistical correction was done using the Benjamini-Hochberg FDR test . Data were analysed using a Confidence interval of 95%. Effect of farm and breed of cows (Gir vs HF cross) were non-significant and removed from further analysis (P < 0.05).