JPH0593712A - Nucleotide sequencer - Google Patents

Abstract

(57) [Summary] [Purpose] To enable automatic determination of the lane in which the input sample migrates with a nucleotide sequencer using a slab gel. [Structure] Until the lane is determined, the fluorescence signals of all coordinates in the scanning direction are stored in the fluorescence time signal storage means 42, and the time variation amount calculation means 46 in the time area set by the time area setting means 44. The time variation of the fluorescence signal is calculated for each measurement coordinate, the cross-correlation function of the calculated time variation and the set slot information is calculated by the cross-correlation function calculating means 72, and the lane determining means 52 calculates the cross-correlation function. Determine the lanes by the amount based on the local maximum of. The base sequence determining means 54 determines the base sequence based on the fluorescence time signal of the determined lane.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for determining the base sequence of a nucleic acid such as DNA, and in particular, a fluorescent labeled nucleic acid fragment is used, for example, in a slab-like gel sampler of a gel electrophoresis apparatus by using, for example, a shark-type comb. Each terminal base is placed in multiple migration lanes that are placed close to each other, and the migration is performed simultaneously.
The present invention relates to an on-line base sequence determination device in which fluorescence is detected by a detection system during migration and a base sequence is determined by a data processing device.

[0002]

2. Description of the Related Art In an on-line type nucleotide sequencer using a slab-like gel, a nucleic acid fragment previously processed by the Sanger method has a terminal base type A (adenine), G (guanine), T (thymine), C. Electrophoresis in separate migration lanes depending on (cytosine). However, the sample introduction position is not fixed when using a sample introduction member such as a shark comb. Such a DNA nucleotide sequencer is disclosed, for example, in Bio / Technology, Vol.6, pp.816-
821 (1988) and J. Biochemical and Biophysical Method.
s, Vol.13, pp.315-323 (1986).
However, those references do not mention how to determine at which coordinate of the slab gel the input sample appears.

[0003]

In an apparatus for determining a base sequence using a slab gel, it is possible to automatically determine which position on the gel is on the coordinate orthogonal to the electrophoretic direction, which is the optimum detection position of the input sample. It's important to make a decision, but traditional devices haven't shown a way around that,
The current situation is that the user manually sets it while watching the detected and output signal pattern. An object of the present invention is to make it possible to automatically determine the lane in which an input sample migrates.

[0004]

FIG. 1 shows the present invention. Four
Reference numeral 2 denotes a fluorescence time signal storage means, which receives a fluorescence signal and scanning data indicating coordinates in the scanning direction, and is orthogonal to the migration direction at least until the lane determination means 52 determines the lane on which the sample migrates ( The fluorescence signal is stored for all the measurement coordinates (in the scanning direction). 44 is a time domain setting means for setting a time domain for lane determination,
Reference numeral 46 is a time variation amount calculating means for calculating the time variation amount of the fluorescence signal, that is, the difference between the maximum value and the minimum value of the fluorescence signal for each measurement coordinate in the set time region, and 50 is the relationship between the sample insertion slot and the sample. Slot information setting means in which the information shown is set, 72 is means for calculating a cross-correlation function between the calculated time variation amount and the set slot information, and 52 is a lane with an amount based on the maximum value of the cross-correlation function. Lane determining means for determining, 54 is a base sequence determining means for determining a base sequence based on the fluorescence time signal of the determined lane.

FIG. 2A shows details of the cross-correlation function calculating means 72 and the lane determining means 52. The cross-correlation function calculating means 72 converts the slot information into a rectangular function 72.
a, a means 72b for synthesizing all the rectangular slot functions, and a calculating means 72c for taking a cross-correlation between the synthesized slot function and the time variation amount from the time variation amount calculating means 46. There is.

The lane determining means 52 is a means 74 for calculating a displacement amount giving a peak of the cross-correlation function, a means 75 for calculating a correspondence between each slot and a lane area based on the displacement amount, and an individual slot. And a means 76 for calculating the point where the time variation becomes maximum. Figure 2
(B) corresponds to claim 2, and the lane determining means 52-1 to 52-1
52-3 respectively correspond to different time regions of migration, the lane determination of those time regions is sequentially executed, and the results are connected by means of a straight line.

[0007]

The slot information rectangularizing means 72a replaces each slot into which a sample has been inserted with a rectangular function having a slot width and a height of 1, and synthesizes it with the means 72b. The cross-correlation function with the signal time fluctuation amount for each lane calculated by the fluctuation amount calculating means 46 is taken, and the displacement amount giving the peak is obtained by the displacement amount calculating means 74 giving the peak of the cross-correlation function. By finding the point where the amount of time variation becomes the maximum in the shifted area, the optimum detection position of the sample can be correctly detected even if the position of the input part shifts to the left or right or there is noise such as glass scratches in the vicinity. I can know. When the rectangular function is formed by the rectangularizing means 72a, the width of the rectangle becomes wider when the samples are placed next to each other in the slots of the shark-shaped comb.

Generally, in gel electrophoresis, there is a phenomenon that the migration position (lane) of the sample slightly shifts as the migration progresses (time elapses) because the salt concentrations of the gel and the sample are not equal (so-called, Lane distortion)
However, as shown in FIG. 1 (C), when the lane determining means is applied to a plurality of time regions of one migration and the lanes obtained by each are connected by a straight line, a peak is always generated even with such distortion. Tracking is possible.

[0009]

EXAMPLE FIG. 3 shows an example of a base sequence determination apparatus to which the present invention is applied. 2 is a slab-like migration gel, and polyacrylamide gel is used. Both ends of the electrophoretic gel 2 are immersed in the electrode layers 4 and 6, and the electrode layers 4 and 6 contain an electrolytic solution. A migration voltage is applied between the electrode layers 4 and 6 by a migration power supply 8. A shark-type comb 33 for injecting a sample is provided at one end of the electrophoretic gel 2. Ten sample introduction slots 34 are formed in the shark type comb 33, and the introduced sample is introduced into each of the terminal bases A, G, C and T using a template.
The sample is four kinds of DNA fragments labeled by a known method with FITC which is a fluorescent substance, and treated by the Sanger method so that each of the bases A, G, T and C comes to the end. FITC is excited by an argon laser with a wavelength of 488 nm and emits fluorescence with a wavelength of 520 nm.

When a migration voltage is applied from the migration power supply 8,
The sample becomes an electrophoretic band 16 and migrates in the electrophoretic gel 2 in the electrophoretic direction 14 over time to be separated, and reaches the measurement section. In the measurement section, an excitation system that irradiates excitation light from an argon laser 18 that emits a laser beam of 488 nm with a condenser lens 20 and a mirror 21, and a migration band 16 if the excitation light beam hits the migration band 16. The fluorescence emitted from the 16 fluorescent substances is collected by the objective lens 22, and the 520 nm interference filter 24 and the condenser lens 26 are collected.
To a photomultiplier tube 28 through the optical fiber bundle 27 and a detection system. In the excitation / detection system including the condenser lens 20, the mirror 21, the objective lens 22, the interference filter 24, the condenser lens 26 and the optical fiber bundle 27, the excitation light beam irradiation position is in a direction orthogonal to the migration direction 14 (scanning direction). 29) A scanning stage 30 which moves mechanically so as to scan on the measurement line for a fixed time, for example, every 5 seconds.
Is provided.

Detection signal of the photomultiplier tube 28 (fluorescence signal)
Is taken into a signal processing microcomputer 31 which is a data processing device through an amplifier and an A / D converter 32. When the excitation light beam irradiates the measurement section on the electrophoretic gel 2, the microcomputer 31 also captures a signal corresponding to the position of the irradiation section as scanning data. In this way, all the fluorescence signals obtained by scanning in the scanning direction are taken into the microcomputer 31 together with the location information.

Next, the operation will be described with reference to FIGS. 4 to 9. The sample introduced by using the shark comb 33 is
As shown in FIG. 4, the template a is used to inject the terminal bases A, G, C, and T in this order, and the template b is used to inject the terminal bases A, G, C, and T in this order. The numbers of the slots 34 into which the respective samples are put are sequentially assigned from the left side as viewed from the optical system of FIG. No sample is put into the 5th and 10th slots. Due to the nature of the shark comb, the sample migrates closely. In addition, the position of the slot 34 is not fixed because the place where the shark-shaped comb 33 is inserted is arbitrary.

The sample is electrophoresed on the gel 2 and the scanning stage 3
An example of the fluorescence pattern obtained by scanning 0 is shown in FIG.
The horizontal axis is coordinates (perpendicular to the migration direction), and in this case, 512 measurement points for a scanning area of 200 mm,
That is, data sampling points are given. The vertical axis represents the detected fluorescence intensity and the depth represents the passage of time.

The first signal 41, which is the signal of the excess primer, is connected to the adjacent lane. Over time, the influence of the tail of the signal of the excess primer decreases, and the signal of the sample becomes clear.

FIG. 6 shows the difference between the maximum value and the minimum value of the fluorescence time signal at each coordinate point (0 to 512) in the time domain 43, that is, the amount of time variation, which is plotted against the coordinates. 6A and 6B, the time domain is (t ₁
˜t ₁ + Δt) and (t ₂ ˜t ₂ + Δt). As a running gel, the thickness is 0.25 mm and the width is 260 m.
As an example of detecting fluorescence at a distance of about 270 mm from the upper end using a 5% polyacrylamide gel having a length of 373 mm and a length of 373 mm, (A) shows 2 hours after the start of migration.
The signal in the time domain up to 40 minutes later, (B) is the signal in the time domain from 6 hours to 6 hours 40 minutes after the start of migration, and signals near 150 bases and 400 bases respectively appear. Position of the vertex x ₁ ~x ₈ of the vertex y ₁ ~y ₈ (B) shown in FIG. 6 (A) is shifted due to the distortion of the lane.

FIG. 7 is a rectangular signal indicating the position of the slot. The horizontal axis is the position of the slot, the width of the rectangle is the width of the slot,
The vertical axis is standardized so that the corresponding part becomes 1. FIG. 7 (1A) is a function representing the A lane of sample a in the time domain (t _{1 to} t ₁ + Δt). Figure 7 (1
B) is a function representing the G lane of sample a. Similarly, although not shown, all lanes of samples a and b, for example, C and T lanes of sample a and A and G of sample b,
For the C and T lanes, slots can be represented by such an isolated rectangular wave. 7 (2A) and (2B) also show rectangular slot functions of the A and G lanes of the sample a, respectively, but they correspond to the time (t _{2 to} t ₂ + Δt) and generally diffuse with the passage of migration time. Since it is possible that the band spreads in the width direction, it is desirable for more precise analysis to change (1A) and (2A) in FIG.

7 (1C) and 7 (2C) respectively show (t ₁
~t ₁ + Δt), is obtained by combining an isolated slot signals of each lane in the time domain determined as described above in _{(t 2 ~t 2 + Δt)} . Below, based on this, as an example, (t
₁ ~t ₁ + _Δt) area shows a calculation to determine the signal lanes A sample a at the. (T ₁ ~t ₁ + Δt) slot information (Fig. 7 (1C)) and variation signal (FIG. 6 (A))
Take the cross-correlation function with. This is shown in FIG. 8 the displacement amount giving a peak in the correlation function of (A) and w _1, again assumed that shifting the individual slot signal in FIG. 7 (1A) only w ₁ taking the product of the signal in FIG. 6 (A). Figure 9
It shows in (A). Coordinates (y ₁ ) in the width direction giving the peak in FIG. 9A of the product of these signals are obtained (t _{1 to} t ₁ + Δ
It is nothing but the coordinates of the lane A of sample a in the area t).

The above calculation is similarly performed in the time domain (t ₂
.About.t ₂ + Δt ₂ ) is performed for lane A of sample a in FIG. 9B. That is, here, in FIG.
The correlation between (B) and FIG. 7 (2C) is obtained (FIG. 8 (B)),
After that, the product of FIG. 7 (2A) and FIG. 6 (B) is obtained. Figure 9
Assuming that the gel coordinate that gives the maximum value in (B) is x ₁ , the lane is y ₁ during the time from time (t ₁ + Δt) to time t _2.
→ It is considered that the peak has moved to x ₁ , so the peak can be traced by approximating this with a straight line. If the above calculation is performed for all the lanes of the samples a and b, that is, the G, C and T lanes of the sample a and the A, G, C and T lanes of the sample b, the determination and tracking of the coordinates of all the lanes are performed. Can be done.

[0019]

According to the present invention, the amount of time variation for each coordinate within the set time domain is calculated, and the coordinate giving the maximum value of this amount of time variation is determined and used as the lane on which the sample migrates. Therefore, even if the shape and position of the comb to which the sample is loaded are indefinite, the optimal detection coordinates are automatically determined as long as the sequence of the loaded samples is known, and automation from the migration after sample loading to the nucleotide sequence determination is possible. .. Furthermore, by performing such lane determination means in a plurality of time regions,
Even if lane distortion occurs, it will be possible to automatically track the lane.

[Brief description of drawings]

FIG. 1 is a block diagram showing the present invention.

2A is a diagram showing in detail a cross-correlation function 72 and a lane determining means 52 in FIG. 1, and FIG. 2B is a diagram showing a portion corresponding to claim 2;

FIG. 3 is a perspective view of an essential part showing an embodiment.

FIG. 4 is a diagram showing an order of sample loading by a template.

FIG. 5 is a diagram showing an example of detected fluorescence signals.

FIG. 6 is a diagram showing the calculated fluorescence time fluctuation amount.

FIG. 7 is a waveform diagram showing a rectangular slot signal.

FIG. 8 is a waveform diagram showing a cross-correlation function between a combination of rectangular slot signals and a time variation amount of a fluorescence signal.

FIG. 9 is a waveform diagram showing a lane determination method.

[Explanation of symbols]

 42 fluorescence time signal storage means 44 time domain setting means 46 time variation calculation means 50 slot information setting means 52, 52-1 to 52-3 lane determination means 54 base sequence determination means 72 cross-correlation function calculation means

Claims (2)

[Claims] 1. An apparatus for determining a base sequence by gel electrophoresis of a fluorescently labeled nucleic acid fragment and sequentially detecting fluorescence during electrophoresis to determine a base sequence, and at least a lane on which a sample migrates by a lane determining means described below. Until is determined, the fluorescence time signal storage means for storing the fluorescence signal for all the measurement coordinates in the direction orthogonal to the migration direction, the time domain setting means for setting the time domain for lane determination, and the set time. A time variation amount of the fluorescence signal for each measurement coordinate in the area, that is, a time variation amount calculating means for calculating the difference between the maximum value and the minimum value of the fluorescence signal, and a slot in which information indicating the relationship between the sample insertion slot and the sample is set. Information setting means, means for calculating the cross-correlation function of the calculated time fluctuation amount and the set slot information, and the maximum value of the cross-correlation function A base sequence determination device comprising: a lane determination unit that determines a lane in an amount based on the amount; and a base sequence determination unit that determines a base sequence based on the fluorescence time signal of the determined lane. 2. The base sequence determination device according to claim 1, wherein a plurality of time regions from the start of migration are taken as regions for calculating the time variation amount, and the lanes calculated in each region are connected by a straight line in a corresponding order.