Using bedtools groupby it should be quite simple, although I don't completely understand the final desired output...

Input example:

cat test.bed
chr1    3000573    3000575    1    2    2    +    G    +    2    2    -    0    0
chr1    3000573    3000575    1    5    5    B    G    +    3    3    -    2    2
chr1    3000573    3000575    1    5    5    B    G    +    4    4    -    1    1
chr1    3000725    3000727    1    1    1    -    G    +    0    0    -    1    1
chr1    3000725    3000727    1    1    1    -    G    +    0    0    -    1    1

Assuming input is sorted by chromosome and position and it has no header:

groupBy -g 1,2,3 \
    -c 5,6,7,8,9,10,11,12,13,14 \
    -o sum,sum,distinct,distinct,distinct,sum,sum,distinct,sum,sum \
    -i test.bed \
| awk 'BEGIN{OFS="\t"} {print $1, $2, $3, $4/$5, $4, $5, $6, $7, $8, $9, $10, $11, $12, $13}'

You get:

chr1    3000573    3000575    1    12    12    +,B    G    +    9    9    -    3    3
chr1    3000725    3000727    1    2    2    -    G    +    0    0    -    2    2

The awk command is just to restore the "ratio" column to be col 5 / col 6.

As it is not clear to me what to do with with columns 7, 8, 9 and 12 I put the option "distinct", which gives a comma separated list of the unique elements found.

Strip the header line and use BEDOPS sort-bed to sort your BED file:

$ tail -n +2 replicates.unsorted.bed | sort-bed - > replicates.bed

Sorting allows you to compare one line to the next and decide how to handle state variables. If your data were not sorted, then you would not be able to process the BED file as a stream.

So, as mentioned, use awk to process your sorted BED file in serial:

$ awk ' \
    BEGIN { \
        FS = "\t"; \
        previousChr = "chrFoo"; \
        previousStart = 0; \
        previousStop = 0; \
        resetFlag = 0; \
    } \
    { \
        split($0, elements, "\t"); \
        currentChr = elements[1]; \
        currentStart = elements[2]; \
        currentStop = elements[3]; \
        if (NR = 1) { \
            currentTotalCSum = elements[5]; \
            currentMethCSum = elements[6]; \
            currentPlusTotalCSum = elements[10]; \
            currentPlusMethCSum = elements[11]; \
            currentMinusTotalCSum = elements[13]; \
            currentMinusMethCSum = elements[14]; \
            resetFlag = 1; \
        } \
        else if (resetFlag == 0) { \
            print previousChr, previousStart, previousStop, currentTotalCSum, currentMethCSum, currentPlusTotalCSum, currentPlusMethCSum, currentMinusTotalCSum, currentMinusMethCSum;
            currentTotalCSum = elements[5]; \
            currentPlusTotalCSum = elements[10]; \
            currentMinusTotalCSum = elements[13]; \
            currentMethCSum = elements[6]; \
            currentPlusMethCSum = elements[11]; \
            currentMinusMethCSum = elements[14]; \
            resetFlag = 1; \
        } \
        else if ((currentChr == previousChr) && (currentStart == previousStart) && (currentStop == previousStop)) { \
            currentTotalCSum += elements[5]; \
            currentPlusTotalCSum += elements[10]; \
            currentMinusTotalCSum += elements[13]; \
            currentMethCSum += elements[5]; \
            currentPlusMethCSum += elements[10]; \
            currentMinusMethCSum += elements[13]; \
        } \
        else { \
            resetFlag = 0; \
        } \
        previousChr = currentChr; \
        previousStart = currentStart; \
        previousStop = currentStop; \
    } \
' replicates.bed > answer.bed

The file answer.bed should have a format close to what you're after, but lacks some columns which you can fill in yourself.

For example, it isn't clear how you handle logic for strand, so I'll leave that to you to add the necessary conditionals.

I also leave out "ratio" and some other columns for brevity, which you can add to the print statement.